{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 46,
   "metadata": {},
   "outputs": [],
   "source": [
    "import tensorflow as tf\n",
    "import matplotlib.pyplot as plt\n",
    "import pandas as pd\n",
    "import imageio.v3 as imageio\n",
    "import numpy as np\n",
    "from tensorflow import keras\n",
    "import tensorflow.keras.layers as layers"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Lab24"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 3.\n",
    "Pakeisti modelių hiperparametrus, optimizavimo funkcijas, tinklo architektūras, aktyvacijos funkcijas."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 61,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model: \"sequential_29\"\n",
      "_________________________________________________________________\n",
      " Layer (type)                Output Shape              Param #   \n",
      "=================================================================\n",
      " dense_41 (Dense)            (None, 20)                15700     \n",
      "                                                                 \n",
      " dense_42 (Dense)            (None, 10)                210       \n",
      "                                                                 \n",
      "=================================================================\n",
      "Total params: 15910 (62.15 KB)\n",
      "Trainable params: 15910 (62.15 KB)\n",
      "Non-trainable params: 0 (0.00 Byte)\n",
      "_________________________________________________________________\n",
      "Epoch 1/200\n",
      "1/1 [==============================] - 2s 2s/step - loss: 0.1158 - sparse_categorical_accuracy: 0.1325 - val_loss: 0.1145 - val_sparse_categorical_accuracy: 0.1317\n",
      "Epoch 2/200\n",
      "1/1 [==============================] - 0s 99ms/step - loss: 0.1152 - sparse_categorical_accuracy: 0.1294 - val_loss: 0.1139 - val_sparse_categorical_accuracy: 0.1297\n",
      "Epoch 3/200\n",
      "1/1 [==============================] - 0s 105ms/step - loss: 0.1146 - sparse_categorical_accuracy: 0.1279 - val_loss: 0.1135 - val_sparse_categorical_accuracy: 0.1270\n",
      "Epoch 4/200\n",
      "1/1 [==============================] - 0s 140ms/step - loss: 0.1142 - sparse_categorical_accuracy: 0.1258 - val_loss: 0.1131 - val_sparse_categorical_accuracy: 0.1266\n",
      "Epoch 5/200\n",
      "1/1 [==============================] - 0s 108ms/step - loss: 0.1138 - sparse_categorical_accuracy: 0.1235 - val_loss: 0.1128 - val_sparse_categorical_accuracy: 0.1255\n",
      "Epoch 6/200\n",
      "1/1 [==============================] - 0s 104ms/step - loss: 0.1135 - sparse_categorical_accuracy: 0.1222 - val_loss: 0.1125 - val_sparse_categorical_accuracy: 0.1228\n",
      "Epoch 7/200\n",
      "1/1 [==============================] - 0s 99ms/step - loss: 0.1133 - sparse_categorical_accuracy: 0.1203 - val_loss: 0.1123 - val_sparse_categorical_accuracy: 0.1217\n",
      "Epoch 8/200\n",
      "1/1 [==============================] - 0s 99ms/step - loss: 0.1131 - sparse_categorical_accuracy: 0.1185 - val_loss: 0.1121 - val_sparse_categorical_accuracy: 0.1203\n",
      "Epoch 9/200\n",
      "1/1 [==============================] - 0s 98ms/step - loss: 0.1129 - sparse_categorical_accuracy: 0.1167 - val_loss: 0.1120 - val_sparse_categorical_accuracy: 0.1182\n",
      "Epoch 10/200\n",
      "1/1 [==============================] - 0s 104ms/step - loss: 0.1128 - sparse_categorical_accuracy: 0.1156 - val_loss: 0.1119 - val_sparse_categorical_accuracy: 0.1181\n",
      "Epoch 11/200\n",
      "1/1 [==============================] - 0s 96ms/step - loss: 0.1126 - sparse_categorical_accuracy: 0.1148 - val_loss: 0.1118 - val_sparse_categorical_accuracy: 0.1166\n",
      "Epoch 12/200\n",
      "1/1 [==============================] - 0s 147ms/step - loss: 0.1125 - sparse_categorical_accuracy: 0.1134 - val_loss: 0.1117 - val_sparse_categorical_accuracy: 0.1161\n",
      "Epoch 13/200\n",
      "1/1 [==============================] - 0s 157ms/step - loss: 0.1124 - sparse_categorical_accuracy: 0.1123 - val_loss: 0.1116 - val_sparse_categorical_accuracy: 0.1150\n",
      "Epoch 14/200\n",
      "1/1 [==============================] - 0s 133ms/step - loss: 0.1124 - sparse_categorical_accuracy: 0.1116 - val_loss: 0.1115 - val_sparse_categorical_accuracy: 0.1143\n",
      "Epoch 15/200\n",
      "1/1 [==============================] - 0s 132ms/step - loss: 0.1123 - sparse_categorical_accuracy: 0.1108 - val_loss: 0.1114 - val_sparse_categorical_accuracy: 0.1141\n",
      "Epoch 16/200\n",
      "1/1 [==============================] - 0s 133ms/step - loss: 0.1122 - sparse_categorical_accuracy: 0.1104 - val_loss: 0.1114 - val_sparse_categorical_accuracy: 0.1136\n",
      "Epoch 17/200\n",
      "1/1 [==============================] - 0s 127ms/step - loss: 0.1121 - sparse_categorical_accuracy: 0.1098 - val_loss: 0.1113 - val_sparse_categorical_accuracy: 0.1132\n",
      "Epoch 18/200\n",
      "1/1 [==============================] - 0s 136ms/step - loss: 0.1121 - sparse_categorical_accuracy: 0.1095 - val_loss: 0.1112 - val_sparse_categorical_accuracy: 0.1126\n",
      "Epoch 19/200\n",
      "1/1 [==============================] - 0s 126ms/step - loss: 0.1120 - sparse_categorical_accuracy: 0.1089 - val_loss: 0.1112 - val_sparse_categorical_accuracy: 0.1127\n",
      "Epoch 20/200\n",
      "1/1 [==============================] - 0s 131ms/step - loss: 0.1120 - sparse_categorical_accuracy: 0.1085 - val_loss: 0.1111 - val_sparse_categorical_accuracy: 0.1127\n",
      "Epoch 21/200\n",
      "1/1 [==============================] - 0s 128ms/step - loss: 0.1119 - sparse_categorical_accuracy: 0.1082 - val_loss: 0.1111 - val_sparse_categorical_accuracy: 0.1119\n",
      "Epoch 22/200\n",
      "1/1 [==============================] - 0s 141ms/step - loss: 0.1118 - sparse_categorical_accuracy: 0.1075 - val_loss: 0.1110 - val_sparse_categorical_accuracy: 0.1108\n",
      "Epoch 23/200\n",
      "1/1 [==============================] - 0s 143ms/step - loss: 0.1118 - sparse_categorical_accuracy: 0.1073 - val_loss: 0.1110 - val_sparse_categorical_accuracy: 0.1098\n",
      "Epoch 24/200\n",
      "1/1 [==============================] - 0s 134ms/step - loss: 0.1117 - sparse_categorical_accuracy: 0.1071 - val_loss: 0.1109 - val_sparse_categorical_accuracy: 0.1096\n",
      "Epoch 25/200\n",
      "1/1 [==============================] - 0s 130ms/step - loss: 0.1117 - sparse_categorical_accuracy: 0.1065 - val_loss: 0.1109 - val_sparse_categorical_accuracy: 0.1085\n",
      "Epoch 26/200\n",
      "1/1 [==============================] - 0s 125ms/step - loss: 0.1117 - sparse_categorical_accuracy: 0.1064 - val_loss: 0.1108 - val_sparse_categorical_accuracy: 0.1079\n",
      "Epoch 27/200\n",
      "1/1 [==============================] - 0s 129ms/step - loss: 0.1116 - sparse_categorical_accuracy: 0.1062 - val_loss: 0.1108 - val_sparse_categorical_accuracy: 0.1078\n",
      "Epoch 28/200\n",
      "1/1 [==============================] - 0s 130ms/step - loss: 0.1116 - sparse_categorical_accuracy: 0.1061 - val_loss: 0.1107 - val_sparse_categorical_accuracy: 0.1078\n",
      "Epoch 29/200\n",
      "1/1 [==============================] - 0s 127ms/step - loss: 0.1115 - sparse_categorical_accuracy: 0.1058 - val_loss: 0.1107 - val_sparse_categorical_accuracy: 0.1074\n",
      "Epoch 30/200\n",
      "1/1 [==============================] - 0s 124ms/step - loss: 0.1115 - sparse_categorical_accuracy: 0.1055 - val_loss: 0.1107 - val_sparse_categorical_accuracy: 0.1066\n",
      "Epoch 31/200\n",
      "1/1 [==============================] - 0s 129ms/step - loss: 0.1114 - sparse_categorical_accuracy: 0.1051 - val_loss: 0.1106 - val_sparse_categorical_accuracy: 0.1067\n",
      "Epoch 32/200\n",
      "1/1 [==============================] - 0s 130ms/step - loss: 0.1114 - sparse_categorical_accuracy: 0.1047 - val_loss: 0.1106 - val_sparse_categorical_accuracy: 0.1071\n",
      "Epoch 33/200\n",
      "1/1 [==============================] - 0s 131ms/step - loss: 0.1114 - sparse_categorical_accuracy: 0.1045 - val_loss: 0.1106 - val_sparse_categorical_accuracy: 0.1072\n",
      "Epoch 34/200\n",
      "1/1 [==============================] - 0s 128ms/step - loss: 0.1113 - sparse_categorical_accuracy: 0.1041 - val_loss: 0.1105 - val_sparse_categorical_accuracy: 0.1068\n",
      "Epoch 35/200\n",
      "1/1 [==============================] - 0s 143ms/step - loss: 0.1113 - sparse_categorical_accuracy: 0.1039 - val_loss: 0.1105 - val_sparse_categorical_accuracy: 0.1064\n",
      "Epoch 36/200\n",
      "1/1 [==============================] - 0s 128ms/step - loss: 0.1113 - sparse_categorical_accuracy: 0.1035 - val_loss: 0.1104 - val_sparse_categorical_accuracy: 0.1061\n",
      "Epoch 37/200\n",
      "1/1 [==============================] - 0s 127ms/step - loss: 0.1112 - sparse_categorical_accuracy: 0.1034 - val_loss: 0.1104 - val_sparse_categorical_accuracy: 0.1067\n",
      "Epoch 38/200\n",
      "1/1 [==============================] - 0s 125ms/step - loss: 0.1112 - sparse_categorical_accuracy: 0.1032 - val_loss: 0.1104 - val_sparse_categorical_accuracy: 0.1065\n",
      "Epoch 39/200\n",
      "1/1 [==============================] - 0s 129ms/step - loss: 0.1112 - sparse_categorical_accuracy: 0.1029 - val_loss: 0.1103 - val_sparse_categorical_accuracy: 0.1060\n",
      "Epoch 40/200\n",
      "1/1 [==============================] - 0s 130ms/step - loss: 0.1111 - sparse_categorical_accuracy: 0.1027 - val_loss: 0.1103 - val_sparse_categorical_accuracy: 0.1057\n",
      "Epoch 41/200\n",
      "1/1 [==============================] - 0s 134ms/step - loss: 0.1111 - sparse_categorical_accuracy: 0.1024 - val_loss: 0.1103 - val_sparse_categorical_accuracy: 0.1050\n",
      "Epoch 42/200\n",
      "1/1 [==============================] - 0s 135ms/step - loss: 0.1111 - sparse_categorical_accuracy: 0.1022 - val_loss: 0.1103 - val_sparse_categorical_accuracy: 0.1051\n",
      "Epoch 43/200\n",
      "1/1 [==============================] - 0s 129ms/step - loss: 0.1110 - sparse_categorical_accuracy: 0.1020 - val_loss: 0.1102 - val_sparse_categorical_accuracy: 0.1043\n",
      "Epoch 44/200\n",
      "1/1 [==============================] - 0s 128ms/step - loss: 0.1110 - sparse_categorical_accuracy: 0.1017 - val_loss: 0.1102 - val_sparse_categorical_accuracy: 0.1037\n",
      "Epoch 45/200\n",
      "1/1 [==============================] - 0s 128ms/step - loss: 0.1110 - sparse_categorical_accuracy: 0.1012 - val_loss: 0.1102 - val_sparse_categorical_accuracy: 0.1026\n",
      "Epoch 46/200\n",
      "1/1 [==============================] - 0s 129ms/step - loss: 0.1109 - sparse_categorical_accuracy: 0.1010 - val_loss: 0.1101 - val_sparse_categorical_accuracy: 0.1027\n",
      "Epoch 47/200\n",
      "1/1 [==============================] - 0s 128ms/step - loss: 0.1109 - sparse_categorical_accuracy: 0.1009 - val_loss: 0.1101 - val_sparse_categorical_accuracy: 0.1020\n",
      "Epoch 48/200\n",
      "1/1 [==============================] - 0s 131ms/step - loss: 0.1109 - sparse_categorical_accuracy: 0.1006 - val_loss: 0.1101 - val_sparse_categorical_accuracy: 0.1017\n",
      "Epoch 49/200\n",
      "1/1 [==============================] - 0s 126ms/step - loss: 0.1109 - sparse_categorical_accuracy: 0.1005 - val_loss: 0.1100 - val_sparse_categorical_accuracy: 0.1018\n",
      "Epoch 50/200\n",
      "1/1 [==============================] - 0s 130ms/step - loss: 0.1108 - sparse_categorical_accuracy: 0.1005 - val_loss: 0.1100 - val_sparse_categorical_accuracy: 0.1013\n",
      "Epoch 51/200\n",
      "1/1 [==============================] - 0s 126ms/step - loss: 0.1108 - sparse_categorical_accuracy: 0.1003 - val_loss: 0.1100 - val_sparse_categorical_accuracy: 0.1015\n",
      "Epoch 52/200\n",
      "1/1 [==============================] - 0s 124ms/step - loss: 0.1108 - sparse_categorical_accuracy: 0.1001 - val_loss: 0.1100 - val_sparse_categorical_accuracy: 0.1014\n",
      "Epoch 53/200\n",
      "1/1 [==============================] - 0s 128ms/step - loss: 0.1107 - sparse_categorical_accuracy: 0.0997 - val_loss: 0.1099 - val_sparse_categorical_accuracy: 0.1018\n",
      "Epoch 54/200\n",
      "1/1 [==============================] - 0s 127ms/step - loss: 0.1107 - sparse_categorical_accuracy: 0.0994 - val_loss: 0.1099 - val_sparse_categorical_accuracy: 0.1016\n",
      "Epoch 55/200\n",
      "1/1 [==============================] - 0s 130ms/step - loss: 0.1107 - sparse_categorical_accuracy: 0.0994 - val_loss: 0.1099 - val_sparse_categorical_accuracy: 0.1017\n",
      "Epoch 56/200\n",
      "1/1 [==============================] - 0s 124ms/step - loss: 0.1107 - sparse_categorical_accuracy: 0.0994 - val_loss: 0.1099 - val_sparse_categorical_accuracy: 0.1014\n",
      "Epoch 57/200\n",
      "1/1 [==============================] - 0s 126ms/step - loss: 0.1106 - sparse_categorical_accuracy: 0.0992 - val_loss: 0.1098 - val_sparse_categorical_accuracy: 0.1012\n",
      "Epoch 58/200\n",
      "1/1 [==============================] - 0s 131ms/step - loss: 0.1106 - sparse_categorical_accuracy: 0.0990 - val_loss: 0.1098 - val_sparse_categorical_accuracy: 0.1010\n",
      "Epoch 59/200\n",
      "1/1 [==============================] - 0s 138ms/step - loss: 0.1106 - sparse_categorical_accuracy: 0.0988 - val_loss: 0.1098 - val_sparse_categorical_accuracy: 0.1007\n",
      "Epoch 60/200\n",
      "1/1 [==============================] - 0s 139ms/step - loss: 0.1106 - sparse_categorical_accuracy: 0.0986 - val_loss: 0.1098 - val_sparse_categorical_accuracy: 0.1009\n",
      "Epoch 61/200\n",
      "1/1 [==============================] - 0s 127ms/step - loss: 0.1105 - sparse_categorical_accuracy: 0.0985 - val_loss: 0.1097 - val_sparse_categorical_accuracy: 0.1004\n",
      "Epoch 62/200\n",
      "1/1 [==============================] - 0s 126ms/step - loss: 0.1105 - sparse_categorical_accuracy: 0.0983 - val_loss: 0.1097 - val_sparse_categorical_accuracy: 0.1001\n",
      "Epoch 63/200\n",
      "1/1 [==============================] - 0s 134ms/step - loss: 0.1105 - sparse_categorical_accuracy: 0.0981 - val_loss: 0.1097 - val_sparse_categorical_accuracy: 0.0993\n",
      "Epoch 64/200\n",
      "1/1 [==============================] - 0s 126ms/step - loss: 0.1105 - sparse_categorical_accuracy: 0.0976 - val_loss: 0.1097 - val_sparse_categorical_accuracy: 0.0991\n",
      "Epoch 65/200\n",
      "1/1 [==============================] - 0s 128ms/step - loss: 0.1104 - sparse_categorical_accuracy: 0.0972 - val_loss: 0.1096 - val_sparse_categorical_accuracy: 0.0991\n",
      "Epoch 66/200\n",
      "1/1 [==============================] - 0s 121ms/step - loss: 0.1104 - sparse_categorical_accuracy: 0.0969 - val_loss: 0.1096 - val_sparse_categorical_accuracy: 0.0989\n",
      "Epoch 67/200\n",
      "1/1 [==============================] - 0s 131ms/step - loss: 0.1104 - sparse_categorical_accuracy: 0.0967 - val_loss: 0.1096 - val_sparse_categorical_accuracy: 0.0981\n",
      "Epoch 68/200\n",
      "1/1 [==============================] - 0s 127ms/step - loss: 0.1104 - sparse_categorical_accuracy: 0.0967 - val_loss: 0.1096 - val_sparse_categorical_accuracy: 0.0981\n",
      "Epoch 69/200\n",
      "1/1 [==============================] - 0s 124ms/step - loss: 0.1103 - sparse_categorical_accuracy: 0.0965 - val_loss: 0.1095 - val_sparse_categorical_accuracy: 0.0981\n",
      "Epoch 70/200\n",
      "1/1 [==============================] - 0s 124ms/step - loss: 0.1103 - sparse_categorical_accuracy: 0.0961 - val_loss: 0.1095 - val_sparse_categorical_accuracy: 0.0977\n",
      "Epoch 71/200\n",
      "1/1 [==============================] - 0s 127ms/step - loss: 0.1103 - sparse_categorical_accuracy: 0.0957 - val_loss: 0.1095 - val_sparse_categorical_accuracy: 0.0973\n",
      "Epoch 72/200\n",
      "1/1 [==============================] - 0s 143ms/step - loss: 0.1103 - sparse_categorical_accuracy: 0.0955 - val_loss: 0.1095 - val_sparse_categorical_accuracy: 0.0972\n",
      "Epoch 73/200\n",
      "1/1 [==============================] - 0s 146ms/step - loss: 0.1102 - sparse_categorical_accuracy: 0.0954 - val_loss: 0.1094 - val_sparse_categorical_accuracy: 0.0972\n",
      "Epoch 74/200\n",
      "1/1 [==============================] - 0s 126ms/step - loss: 0.1102 - sparse_categorical_accuracy: 0.0951 - val_loss: 0.1094 - val_sparse_categorical_accuracy: 0.0977\n",
      "Epoch 75/200\n",
      "1/1 [==============================] - 0s 126ms/step - loss: 0.1102 - sparse_categorical_accuracy: 0.0948 - val_loss: 0.1094 - val_sparse_categorical_accuracy: 0.0975\n",
      "Epoch 76/200\n",
      "1/1 [==============================] - 0s 134ms/step - loss: 0.1102 - sparse_categorical_accuracy: 0.0948 - val_loss: 0.1094 - val_sparse_categorical_accuracy: 0.0976\n",
      "Epoch 77/200\n",
      "1/1 [==============================] - 0s 131ms/step - loss: 0.1101 - sparse_categorical_accuracy: 0.0946 - val_loss: 0.1094 - val_sparse_categorical_accuracy: 0.0974\n",
      "Epoch 78/200\n",
      "1/1 [==============================] - 0s 135ms/step - loss: 0.1101 - sparse_categorical_accuracy: 0.0945 - val_loss: 0.1093 - val_sparse_categorical_accuracy: 0.0970\n",
      "Epoch 79/200\n",
      "1/1 [==============================] - 0s 130ms/step - loss: 0.1101 - sparse_categorical_accuracy: 0.0946 - val_loss: 0.1093 - val_sparse_categorical_accuracy: 0.0968\n",
      "Epoch 80/200\n",
      "1/1 [==============================] - 0s 140ms/step - loss: 0.1101 - sparse_categorical_accuracy: 0.0946 - val_loss: 0.1093 - val_sparse_categorical_accuracy: 0.0967\n",
      "Epoch 81/200\n",
      "1/1 [==============================] - 0s 142ms/step - loss: 0.1101 - sparse_categorical_accuracy: 0.0943 - val_loss: 0.1093 - val_sparse_categorical_accuracy: 0.0964\n",
      "Epoch 82/200\n",
      "1/1 [==============================] - 0s 128ms/step - loss: 0.1100 - sparse_categorical_accuracy: 0.0938 - val_loss: 0.1093 - val_sparse_categorical_accuracy: 0.0958\n",
      "Epoch 83/200\n",
      "1/1 [==============================] - 0s 124ms/step - loss: 0.1100 - sparse_categorical_accuracy: 0.0934 - val_loss: 0.1092 - val_sparse_categorical_accuracy: 0.0957\n",
      "Epoch 84/200\n",
      "1/1 [==============================] - 0s 123ms/step - loss: 0.1100 - sparse_categorical_accuracy: 0.0931 - val_loss: 0.1092 - val_sparse_categorical_accuracy: 0.0948\n",
      "Epoch 85/200\n",
      "1/1 [==============================] - 0s 127ms/step - loss: 0.1100 - sparse_categorical_accuracy: 0.0929 - val_loss: 0.1092 - val_sparse_categorical_accuracy: 0.0945\n",
      "Epoch 86/200\n",
      "1/1 [==============================] - 0s 127ms/step - loss: 0.1100 - sparse_categorical_accuracy: 0.0926 - val_loss: 0.1092 - val_sparse_categorical_accuracy: 0.0947\n",
      "Epoch 87/200\n",
      "1/1 [==============================] - 0s 127ms/step - loss: 0.1099 - sparse_categorical_accuracy: 0.0922 - val_loss: 0.1091 - val_sparse_categorical_accuracy: 0.0943\n",
      "Epoch 88/200\n",
      "1/1 [==============================] - 0s 127ms/step - loss: 0.1099 - sparse_categorical_accuracy: 0.0922 - val_loss: 0.1091 - val_sparse_categorical_accuracy: 0.0945\n",
      "Epoch 89/200\n",
      "1/1 [==============================] - 0s 124ms/step - loss: 0.1099 - sparse_categorical_accuracy: 0.0923 - val_loss: 0.1091 - val_sparse_categorical_accuracy: 0.0946\n",
      "Epoch 90/200\n",
      "1/1 [==============================] - 0s 125ms/step - loss: 0.1099 - sparse_categorical_accuracy: 0.0920 - val_loss: 0.1091 - val_sparse_categorical_accuracy: 0.0947\n",
      "Epoch 91/200\n",
      "1/1 [==============================] - 0s 124ms/step - loss: 0.1099 - sparse_categorical_accuracy: 0.0921 - val_loss: 0.1091 - val_sparse_categorical_accuracy: 0.0931\n",
      "Epoch 92/200\n",
      "1/1 [==============================] - 0s 127ms/step - loss: 0.1098 - sparse_categorical_accuracy: 0.0921 - val_loss: 0.1091 - val_sparse_categorical_accuracy: 0.0927\n",
      "Epoch 93/200\n",
      "1/1 [==============================] - 0s 128ms/step - loss: 0.1098 - sparse_categorical_accuracy: 0.0921 - val_loss: 0.1090 - val_sparse_categorical_accuracy: 0.0920\n",
      "Epoch 94/200\n",
      "1/1 [==============================] - 0s 133ms/step - loss: 0.1098 - sparse_categorical_accuracy: 0.0919 - val_loss: 0.1090 - val_sparse_categorical_accuracy: 0.0925\n",
      "Epoch 95/200\n",
      "1/1 [==============================] - 0s 139ms/step - loss: 0.1098 - sparse_categorical_accuracy: 0.0918 - val_loss: 0.1090 - val_sparse_categorical_accuracy: 0.0927\n",
      "Epoch 96/200\n",
      "1/1 [==============================] - 0s 134ms/step - loss: 0.1098 - sparse_categorical_accuracy: 0.0919 - val_loss: 0.1090 - val_sparse_categorical_accuracy: 0.0929\n",
      "Epoch 97/200\n",
      "1/1 [==============================] - 0s 127ms/step - loss: 0.1097 - sparse_categorical_accuracy: 0.0917 - val_loss: 0.1090 - val_sparse_categorical_accuracy: 0.0932\n",
      "Epoch 98/200\n",
      "1/1 [==============================] - 0s 133ms/step - loss: 0.1097 - sparse_categorical_accuracy: 0.0916 - val_loss: 0.1090 - val_sparse_categorical_accuracy: 0.0930\n",
      "Epoch 99/200\n",
      "1/1 [==============================] - 0s 127ms/step - loss: 0.1097 - sparse_categorical_accuracy: 0.0913 - val_loss: 0.1089 - val_sparse_categorical_accuracy: 0.0927\n",
      "Epoch 100/200\n",
      "1/1 [==============================] - 0s 127ms/step - loss: 0.1097 - sparse_categorical_accuracy: 0.0912 - val_loss: 0.1089 - val_sparse_categorical_accuracy: 0.0929\n",
      "Epoch 101/200\n",
      "1/1 [==============================] - 0s 127ms/step - loss: 0.1097 - sparse_categorical_accuracy: 0.0909 - val_loss: 0.1089 - val_sparse_categorical_accuracy: 0.0927\n",
      "Epoch 102/200\n",
      "1/1 [==============================] - 0s 137ms/step - loss: 0.1097 - sparse_categorical_accuracy: 0.0910 - val_loss: 0.1089 - val_sparse_categorical_accuracy: 0.0928\n",
      "Epoch 103/200\n",
      "1/1 [==============================] - 0s 126ms/step - loss: 0.1096 - sparse_categorical_accuracy: 0.0908 - val_loss: 0.1089 - val_sparse_categorical_accuracy: 0.0931\n",
      "Epoch 104/200\n",
      "1/1 [==============================] - 0s 128ms/step - loss: 0.1096 - sparse_categorical_accuracy: 0.0906 - val_loss: 0.1089 - val_sparse_categorical_accuracy: 0.0927\n",
      "Epoch 105/200\n",
      "1/1 [==============================] - 0s 128ms/step - loss: 0.1096 - sparse_categorical_accuracy: 0.0903 - val_loss: 0.1088 - val_sparse_categorical_accuracy: 0.0933\n",
      "Epoch 106/200\n",
      "1/1 [==============================] - 0s 123ms/step - loss: 0.1096 - sparse_categorical_accuracy: 0.0905 - val_loss: 0.1088 - val_sparse_categorical_accuracy: 0.0924\n",
      "Epoch 107/200\n",
      "1/1 [==============================] - 0s 125ms/step - loss: 0.1096 - sparse_categorical_accuracy: 0.0900 - val_loss: 0.1088 - val_sparse_categorical_accuracy: 0.0930\n",
      "Epoch 108/200\n",
      "1/1 [==============================] - 0s 126ms/step - loss: 0.1096 - sparse_categorical_accuracy: 0.0901 - val_loss: 0.1088 - val_sparse_categorical_accuracy: 0.0921\n",
      "Epoch 109/200\n",
      "1/1 [==============================] - 0s 124ms/step - loss: 0.1096 - sparse_categorical_accuracy: 0.0897 - val_loss: 0.1088 - val_sparse_categorical_accuracy: 0.0921\n",
      "Epoch 110/200\n",
      "1/1 [==============================] - 0s 126ms/step - loss: 0.1095 - sparse_categorical_accuracy: 0.0900 - val_loss: 0.1088 - val_sparse_categorical_accuracy: 0.0904\n",
      "Epoch 111/200\n",
      "1/1 [==============================] - 0s 121ms/step - loss: 0.1095 - sparse_categorical_accuracy: 0.0896 - val_loss: 0.1088 - val_sparse_categorical_accuracy: 0.0916\n",
      "Epoch 112/200\n",
      "1/1 [==============================] - 0s 131ms/step - loss: 0.1095 - sparse_categorical_accuracy: 0.0899 - val_loss: 0.1088 - val_sparse_categorical_accuracy: 0.0902\n",
      "Epoch 113/200\n",
      "1/1 [==============================] - 0s 135ms/step - loss: 0.1095 - sparse_categorical_accuracy: 0.0892 - val_loss: 0.1087 - val_sparse_categorical_accuracy: 0.0911\n",
      "Epoch 114/200\n",
      "1/1 [==============================] - 0s 136ms/step - loss: 0.1095 - sparse_categorical_accuracy: 0.0894 - val_loss: 0.1087 - val_sparse_categorical_accuracy: 0.0898\n",
      "Epoch 115/200\n",
      "1/1 [==============================] - 0s 126ms/step - loss: 0.1095 - sparse_categorical_accuracy: 0.0889 - val_loss: 0.1087 - val_sparse_categorical_accuracy: 0.0907\n",
      "Epoch 116/200\n",
      "1/1 [==============================] - 0s 146ms/step - loss: 0.1095 - sparse_categorical_accuracy: 0.0894 - val_loss: 0.1087 - val_sparse_categorical_accuracy: 0.0893\n",
      "Epoch 117/200\n",
      "1/1 [==============================] - 0s 148ms/step - loss: 0.1095 - sparse_categorical_accuracy: 0.0889 - val_loss: 0.1087 - val_sparse_categorical_accuracy: 0.0889\n",
      "Epoch 118/200\n",
      "1/1 [==============================] - 0s 125ms/step - loss: 0.1095 - sparse_categorical_accuracy: 0.0891 - val_loss: 0.1087 - val_sparse_categorical_accuracy: 0.0901\n",
      "Epoch 119/200\n",
      "1/1 [==============================] - 0s 130ms/step - loss: 0.1094 - sparse_categorical_accuracy: 0.0894 - val_loss: 0.1087 - val_sparse_categorical_accuracy: 0.0887\n",
      "Epoch 120/200\n",
      "1/1 [==============================] - 0s 125ms/step - loss: 0.1094 - sparse_categorical_accuracy: 0.0889 - val_loss: 0.1087 - val_sparse_categorical_accuracy: 0.0899\n",
      "Epoch 121/200\n",
      "1/1 [==============================] - 0s 125ms/step - loss: 0.1094 - sparse_categorical_accuracy: 0.0893 - val_loss: 0.1087 - val_sparse_categorical_accuracy: 0.0885\n",
      "Epoch 122/200\n",
      "1/1 [==============================] - 0s 128ms/step - loss: 0.1094 - sparse_categorical_accuracy: 0.0887 - val_loss: 0.1086 - val_sparse_categorical_accuracy: 0.0893\n",
      "Epoch 123/200\n",
      "1/1 [==============================] - 0s 128ms/step - loss: 0.1094 - sparse_categorical_accuracy: 0.0893 - val_loss: 0.1086 - val_sparse_categorical_accuracy: 0.0890\n",
      "Epoch 124/200\n",
      "1/1 [==============================] - 0s 126ms/step - loss: 0.1094 - sparse_categorical_accuracy: 0.0886 - val_loss: 0.1086 - val_sparse_categorical_accuracy: 0.0894\n",
      "Epoch 125/200\n",
      "1/1 [==============================] - 0s 126ms/step - loss: 0.1094 - sparse_categorical_accuracy: 0.0889 - val_loss: 0.1086 - val_sparse_categorical_accuracy: 0.0883\n",
      "Epoch 126/200\n",
      "1/1 [==============================] - 0s 123ms/step - loss: 0.1094 - sparse_categorical_accuracy: 0.0880 - val_loss: 0.1086 - val_sparse_categorical_accuracy: 0.0898\n",
      "Epoch 127/200\n",
      "1/1 [==============================] - 0s 130ms/step - loss: 0.1094 - sparse_categorical_accuracy: 0.0887 - val_loss: 0.1086 - val_sparse_categorical_accuracy: 0.0887\n",
      "Epoch 128/200\n",
      "1/1 [==============================] - 0s 126ms/step - loss: 0.1093 - sparse_categorical_accuracy: 0.0881 - val_loss: 0.1086 - val_sparse_categorical_accuracy: 0.0892\n",
      "Epoch 129/200\n",
      "1/1 [==============================] - 0s 128ms/step - loss: 0.1093 - sparse_categorical_accuracy: 0.0887 - val_loss: 0.1086 - val_sparse_categorical_accuracy: 0.0880\n",
      "Epoch 130/200\n",
      "1/1 [==============================] - 0s 132ms/step - loss: 0.1093 - sparse_categorical_accuracy: 0.0880 - val_loss: 0.1086 - val_sparse_categorical_accuracy: 0.0882\n",
      "Epoch 131/200\n",
      "1/1 [==============================] - 0s 130ms/step - loss: 0.1093 - sparse_categorical_accuracy: 0.0880 - val_loss: 0.1086 - val_sparse_categorical_accuracy: 0.0898\n",
      "Epoch 132/200\n",
      "1/1 [==============================] - 0s 134ms/step - loss: 0.1093 - sparse_categorical_accuracy: 0.0885 - val_loss: 0.1085 - val_sparse_categorical_accuracy: 0.0884\n",
      "Epoch 133/200\n",
      "1/1 [==============================] - 0s 125ms/step - loss: 0.1093 - sparse_categorical_accuracy: 0.0875 - val_loss: 0.1085 - val_sparse_categorical_accuracy: 0.0891\n",
      "Epoch 134/200\n",
      "1/1 [==============================] - 0s 123ms/step - loss: 0.1093 - sparse_categorical_accuracy: 0.0884 - val_loss: 0.1085 - val_sparse_categorical_accuracy: 0.0878\n",
      "Epoch 135/200\n",
      "1/1 [==============================] - 0s 128ms/step - loss: 0.1093 - sparse_categorical_accuracy: 0.0876 - val_loss: 0.1085 - val_sparse_categorical_accuracy: 0.0889\n",
      "Epoch 136/200\n",
      "1/1 [==============================] - 0s 130ms/step - loss: 0.1093 - sparse_categorical_accuracy: 0.0884 - val_loss: 0.1085 - val_sparse_categorical_accuracy: 0.0881\n",
      "Epoch 137/200\n",
      "1/1 [==============================] - 0s 135ms/step - loss: 0.1093 - sparse_categorical_accuracy: 0.0875 - val_loss: 0.1085 - val_sparse_categorical_accuracy: 0.0888\n",
      "Epoch 138/200\n",
      "1/1 [==============================] - 0s 125ms/step - loss: 0.1093 - sparse_categorical_accuracy: 0.0884 - val_loss: 0.1085 - val_sparse_categorical_accuracy: 0.0880\n",
      "Epoch 139/200\n",
      "1/1 [==============================] - 0s 127ms/step - loss: 0.1093 - sparse_categorical_accuracy: 0.0874 - val_loss: 0.1085 - val_sparse_categorical_accuracy: 0.0890\n",
      "Epoch 140/200\n",
      "1/1 [==============================] - 0s 130ms/step - loss: 0.1092 - sparse_categorical_accuracy: 0.0882 - val_loss: 0.1085 - val_sparse_categorical_accuracy: 0.0876\n",
      "Epoch 141/200\n",
      "1/1 [==============================] - 0s 129ms/step - loss: 0.1092 - sparse_categorical_accuracy: 0.0871 - val_loss: 0.1085 - val_sparse_categorical_accuracy: 0.0890\n",
      "Epoch 142/200\n",
      "1/1 [==============================] - 0s 123ms/step - loss: 0.1092 - sparse_categorical_accuracy: 0.0876 - val_loss: 0.1085 - val_sparse_categorical_accuracy: 0.0877\n",
      "Epoch 143/200\n",
      "1/1 [==============================] - 0s 128ms/step - loss: 0.1092 - sparse_categorical_accuracy: 0.0867 - val_loss: 0.1085 - val_sparse_categorical_accuracy: 0.0890\n",
      "Epoch 144/200\n",
      "1/1 [==============================] - 0s 124ms/step - loss: 0.1092 - sparse_categorical_accuracy: 0.0875 - val_loss: 0.1085 - val_sparse_categorical_accuracy: 0.0879\n",
      "Epoch 145/200\n",
      "1/1 [==============================] - 0s 128ms/step - loss: 0.1092 - sparse_categorical_accuracy: 0.0863 - val_loss: 0.1084 - val_sparse_categorical_accuracy: 0.0893\n",
      "Epoch 146/200\n",
      "1/1 [==============================] - 0s 121ms/step - loss: 0.1092 - sparse_categorical_accuracy: 0.0872 - val_loss: 0.1084 - val_sparse_categorical_accuracy: 0.0883\n",
      "Epoch 147/200\n",
      "1/1 [==============================] - 0s 126ms/step - loss: 0.1092 - sparse_categorical_accuracy: 0.0861 - val_loss: 0.1084 - val_sparse_categorical_accuracy: 0.0895\n",
      "Epoch 148/200\n",
      "1/1 [==============================] - 0s 131ms/step - loss: 0.1092 - sparse_categorical_accuracy: 0.0869 - val_loss: 0.1084 - val_sparse_categorical_accuracy: 0.0886\n",
      "Epoch 149/200\n",
      "1/1 [==============================] - 0s 129ms/step - loss: 0.1092 - sparse_categorical_accuracy: 0.0860 - val_loss: 0.1084 - val_sparse_categorical_accuracy: 0.0893\n",
      "Epoch 150/200\n",
      "1/1 [==============================] - 0s 133ms/step - loss: 0.1092 - sparse_categorical_accuracy: 0.0868 - val_loss: 0.1084 - val_sparse_categorical_accuracy: 0.0880\n",
      "Epoch 151/200\n",
      "1/1 [==============================] - 0s 128ms/step - loss: 0.1092 - sparse_categorical_accuracy: 0.0859 - val_loss: 0.1084 - val_sparse_categorical_accuracy: 0.0891\n",
      "Epoch 152/200\n",
      "1/1 [==============================] - 0s 137ms/step - loss: 0.1092 - sparse_categorical_accuracy: 0.0866 - val_loss: 0.1084 - val_sparse_categorical_accuracy: 0.0883\n",
      "Epoch 153/200\n",
      "1/1 [==============================] - 0s 157ms/step - loss: 0.1092 - sparse_categorical_accuracy: 0.0856 - val_loss: 0.1084 - val_sparse_categorical_accuracy: 0.0892\n",
      "Epoch 154/200\n",
      "1/1 [==============================] - 0s 137ms/step - loss: 0.1091 - sparse_categorical_accuracy: 0.0865 - val_loss: 0.1084 - val_sparse_categorical_accuracy: 0.0883\n",
      "Epoch 155/200\n",
      "1/1 [==============================] - 0s 137ms/step - loss: 0.1091 - sparse_categorical_accuracy: 0.0854 - val_loss: 0.1084 - val_sparse_categorical_accuracy: 0.0892\n",
      "Epoch 156/200\n",
      "1/1 [==============================] - 0s 135ms/step - loss: 0.1091 - sparse_categorical_accuracy: 0.0865 - val_loss: 0.1084 - val_sparse_categorical_accuracy: 0.0883\n",
      "Epoch 157/200\n",
      "1/1 [==============================] - 0s 137ms/step - loss: 0.1091 - sparse_categorical_accuracy: 0.0853 - val_loss: 0.1084 - val_sparse_categorical_accuracy: 0.0864\n",
      "Epoch 158/200\n",
      "1/1 [==============================] - 0s 141ms/step - loss: 0.1091 - sparse_categorical_accuracy: 0.0840 - val_loss: 0.1084 - val_sparse_categorical_accuracy: 0.0890\n",
      "Epoch 159/200\n",
      "1/1 [==============================] - 0s 135ms/step - loss: 0.1091 - sparse_categorical_accuracy: 0.0858 - val_loss: 0.1084 - val_sparse_categorical_accuracy: 0.0874\n",
      "Epoch 160/200\n",
      "1/1 [==============================] - 0s 132ms/step - loss: 0.1091 - sparse_categorical_accuracy: 0.0846 - val_loss: 0.1084 - val_sparse_categorical_accuracy: 0.0860\n",
      "Epoch 161/200\n",
      "1/1 [==============================] - 0s 124ms/step - loss: 0.1091 - sparse_categorical_accuracy: 0.0836 - val_loss: 0.1083 - val_sparse_categorical_accuracy: 0.0881\n",
      "Epoch 162/200\n",
      "1/1 [==============================] - 0s 131ms/step - loss: 0.1091 - sparse_categorical_accuracy: 0.0856 - val_loss: 0.1083 - val_sparse_categorical_accuracy: 0.0866\n",
      "Epoch 163/200\n",
      "1/1 [==============================] - 0s 125ms/step - loss: 0.1091 - sparse_categorical_accuracy: 0.0844 - val_loss: 0.1083 - val_sparse_categorical_accuracy: 0.0875\n",
      "Epoch 164/200\n",
      "1/1 [==============================] - 0s 125ms/step - loss: 0.1091 - sparse_categorical_accuracy: 0.0856 - val_loss: 0.1083 - val_sparse_categorical_accuracy: 0.0862\n",
      "Epoch 165/200\n",
      "1/1 [==============================] - 0s 139ms/step - loss: 0.1091 - sparse_categorical_accuracy: 0.0843 - val_loss: 0.1083 - val_sparse_categorical_accuracy: 0.0840\n",
      "Epoch 166/200\n",
      "1/1 [==============================] - 0s 137ms/step - loss: 0.1091 - sparse_categorical_accuracy: 0.0831 - val_loss: 0.1083 - val_sparse_categorical_accuracy: 0.0877\n",
      "Epoch 167/200\n",
      "1/1 [==============================] - 0s 132ms/step - loss: 0.1091 - sparse_categorical_accuracy: 0.0853 - val_loss: 0.1083 - val_sparse_categorical_accuracy: 0.0856\n",
      "Epoch 168/200\n",
      "1/1 [==============================] - 0s 134ms/step - loss: 0.1091 - sparse_categorical_accuracy: 0.0842 - val_loss: 0.1083 - val_sparse_categorical_accuracy: 0.0840\n",
      "Epoch 169/200\n",
      "1/1 [==============================] - 0s 127ms/step - loss: 0.1091 - sparse_categorical_accuracy: 0.0829 - val_loss: 0.1083 - val_sparse_categorical_accuracy: 0.0863\n",
      "Epoch 170/200\n",
      "1/1 [==============================] - 0s 129ms/step - loss: 0.1091 - sparse_categorical_accuracy: 0.0842 - val_loss: 0.1083 - val_sparse_categorical_accuracy: 0.0882\n",
      "Epoch 171/200\n",
      "1/1 [==============================] - 0s 125ms/step - loss: 0.1091 - sparse_categorical_accuracy: 0.0853 - val_loss: 0.1083 - val_sparse_categorical_accuracy: 0.0837\n",
      "Epoch 172/200\n",
      "1/1 [==============================] - 0s 155ms/step - loss: 0.1090 - sparse_categorical_accuracy: 0.0826 - val_loss: 0.1083 - val_sparse_categorical_accuracy: 0.0867\n",
      "Epoch 173/200\n",
      "1/1 [==============================] - 0s 128ms/step - loss: 0.1090 - sparse_categorical_accuracy: 0.0841 - val_loss: 0.1083 - val_sparse_categorical_accuracy: 0.0880\n",
      "Epoch 174/200\n",
      "1/1 [==============================] - 0s 140ms/step - loss: 0.1090 - sparse_categorical_accuracy: 0.0852 - val_loss: 0.1083 - val_sparse_categorical_accuracy: 0.0841\n",
      "Epoch 175/200\n",
      "1/1 [==============================] - 0s 126ms/step - loss: 0.1090 - sparse_categorical_accuracy: 0.0825 - val_loss: 0.1083 - val_sparse_categorical_accuracy: 0.0866\n",
      "Epoch 176/200\n",
      "1/1 [==============================] - 0s 127ms/step - loss: 0.1090 - sparse_categorical_accuracy: 0.0840 - val_loss: 0.1083 - val_sparse_categorical_accuracy: 0.0884\n",
      "Epoch 177/200\n",
      "1/1 [==============================] - 0s 124ms/step - loss: 0.1090 - sparse_categorical_accuracy: 0.0855 - val_loss: 0.1083 - val_sparse_categorical_accuracy: 0.0846\n",
      "Epoch 178/200\n",
      "1/1 [==============================] - 0s 125ms/step - loss: 0.1090 - sparse_categorical_accuracy: 0.0824 - val_loss: 0.1083 - val_sparse_categorical_accuracy: 0.0870\n",
      "Epoch 179/200\n",
      "1/1 [==============================] - 0s 126ms/step - loss: 0.1090 - sparse_categorical_accuracy: 0.0840 - val_loss: 0.1083 - val_sparse_categorical_accuracy: 0.0849\n",
      "Epoch 180/200\n",
      "1/1 [==============================] - 0s 124ms/step - loss: 0.1090 - sparse_categorical_accuracy: 0.0829 - val_loss: 0.1083 - val_sparse_categorical_accuracy: 0.0894\n",
      "Epoch 181/200\n",
      "1/1 [==============================] - 0s 126ms/step - loss: 0.1090 - sparse_categorical_accuracy: 0.0854 - val_loss: 0.1082 - val_sparse_categorical_accuracy: 0.0873\n",
      "Epoch 182/200\n",
      "1/1 [==============================] - 0s 126ms/step - loss: 0.1090 - sparse_categorical_accuracy: 0.0840 - val_loss: 0.1082 - val_sparse_categorical_accuracy: 0.0856\n",
      "Epoch 183/200\n",
      "1/1 [==============================] - 0s 125ms/step - loss: 0.1090 - sparse_categorical_accuracy: 0.0838 - val_loss: 0.1082 - val_sparse_categorical_accuracy: 0.0841\n",
      "Epoch 184/200\n",
      "1/1 [==============================] - 0s 146ms/step - loss: 0.1090 - sparse_categorical_accuracy: 0.0819 - val_loss: 0.1082 - val_sparse_categorical_accuracy: 0.0867\n",
      "Epoch 185/200\n",
      "1/1 [==============================] - 0s 152ms/step - loss: 0.1090 - sparse_categorical_accuracy: 0.0838 - val_loss: 0.1082 - val_sparse_categorical_accuracy: 0.0886\n",
      "Epoch 186/200\n",
      "1/1 [==============================] - 0s 133ms/step - loss: 0.1090 - sparse_categorical_accuracy: 0.0850 - val_loss: 0.1082 - val_sparse_categorical_accuracy: 0.0843\n",
      "Epoch 187/200\n",
      "1/1 [==============================] - 0s 129ms/step - loss: 0.1090 - sparse_categorical_accuracy: 0.0815 - val_loss: 0.1082 - val_sparse_categorical_accuracy: 0.0854\n",
      "Epoch 188/200\n",
      "1/1 [==============================] - 0s 127ms/step - loss: 0.1090 - sparse_categorical_accuracy: 0.0828 - val_loss: 0.1082 - val_sparse_categorical_accuracy: 0.0864\n",
      "Epoch 189/200\n",
      "1/1 [==============================] - 0s 127ms/step - loss: 0.1090 - sparse_categorical_accuracy: 0.0834 - val_loss: 0.1082 - val_sparse_categorical_accuracy: 0.0876\n",
      "Epoch 190/200\n",
      "1/1 [==============================] - 0s 128ms/step - loss: 0.1090 - sparse_categorical_accuracy: 0.0847 - val_loss: 0.1082 - val_sparse_categorical_accuracy: 0.0833\n",
      "Epoch 191/200\n",
      "1/1 [==============================] - 0s 139ms/step - loss: 0.1090 - sparse_categorical_accuracy: 0.0811 - val_loss: 0.1082 - val_sparse_categorical_accuracy: 0.0843\n",
      "Epoch 192/200\n",
      "1/1 [==============================] - 0s 128ms/step - loss: 0.1090 - sparse_categorical_accuracy: 0.0817 - val_loss: 0.1082 - val_sparse_categorical_accuracy: 0.0859\n",
      "Epoch 193/200\n",
      "1/1 [==============================] - 0s 127ms/step - loss: 0.1090 - sparse_categorical_accuracy: 0.0829 - val_loss: 0.1082 - val_sparse_categorical_accuracy: 0.0867\n",
      "Epoch 194/200\n",
      "1/1 [==============================] - 0s 125ms/step - loss: 0.1090 - sparse_categorical_accuracy: 0.0834 - val_loss: 0.1082 - val_sparse_categorical_accuracy: 0.0836\n",
      "Epoch 195/200\n",
      "1/1 [==============================] - 0s 126ms/step - loss: 0.1090 - sparse_categorical_accuracy: 0.0808 - val_loss: 0.1082 - val_sparse_categorical_accuracy: 0.0877\n",
      "Epoch 196/200\n",
      "1/1 [==============================] - 0s 126ms/step - loss: 0.1090 - sparse_categorical_accuracy: 0.0844 - val_loss: 0.1082 - val_sparse_categorical_accuracy: 0.0850\n",
      "Epoch 197/200\n",
      "1/1 [==============================] - 0s 129ms/step - loss: 0.1089 - sparse_categorical_accuracy: 0.0822 - val_loss: 0.1082 - val_sparse_categorical_accuracy: 0.0863\n",
      "Epoch 198/200\n",
      "1/1 [==============================] - 0s 98ms/step - loss: 0.1089 - sparse_categorical_accuracy: 0.0828 - val_loss: 0.1082 - val_sparse_categorical_accuracy: 0.0833\n",
      "Epoch 199/200\n",
      "1/1 [==============================] - 0s 92ms/step - loss: 0.1089 - sparse_categorical_accuracy: 0.0800 - val_loss: 0.1082 - val_sparse_categorical_accuracy: 0.0881\n",
      "Epoch 200/200\n",
      "1/1 [==============================] - 0s 99ms/step - loss: 0.1089 - sparse_categorical_accuracy: 0.0841 - val_loss: 0.1082 - val_sparse_categorical_accuracy: 0.0842\n"
     ]
    }
   ],
   "source": [
    "# train mnist\n",
    "\n",
    "mnist = tf.keras.datasets.mnist\n",
    "(train_images0, train_labels0), (test_images0, test_labels0) = mnist.load_data()\n",
    "\n",
    "test_images  = test_images0.reshape(10000, 784)\n",
    "train_images = train_images0.reshape(60000, 784)\n",
    "\n",
    "test_images  = test_images/255.0\n",
    "train_images = train_images/255.0\n",
    "\n",
    "keras_model = tf.keras.models.Sequential([\n",
    "  tf.keras.layers.Dense(20, activation='relu'),\n",
    "  tf.keras.layers.Dense(10, activation='softmax')\n",
    "])\n",
    "\n",
    "keras_model.build(input_shape=[None,784])\n",
    "\n",
    "keras_model.summary()\n",
    "\n",
    "keras_model.compile(\n",
    "  optimizer=tf.keras.optimizers.SGD(0.2),\n",
    "  loss=tf.keras.losses.CategoricalHinge(),\n",
    "  metrics=[tf.keras.metrics.SparseCategoricalAccuracy()],\n",
    ")\n",
    "\n",
    "# Train loop\n",
    "history = keras_model.fit(\n",
    "  train_images,\n",
    "  train_labels0,\n",
    "  batch_size=len(train_images),\n",
    "  epochs=200,\n",
    "  validation_data=(test_images, test_labels0),\n",
    ")\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 62,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Key: loss\n",
      "Key: sparse_categorical_accuracy\n",
      "Key: val_loss\n",
      "Key: val_sparse_categorical_accuracy\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# vizualize mnist\n",
    "\n",
    "for item in history.history:\n",
    "    print(\"Key:\",item)\n",
    "\n",
    "plt.plot(history.history['loss'],label=\"train\")\n",
    "plt.plot(history.history['val_loss'],label=\"validation\")\n",
    "plt.title('Model Loss')\n",
    "plt.yscale('log')\n",
    "plt.ylabel('Cross Entropy')\n",
    "plt.xlabel('Iteration')\n",
    "plt.grid()\n",
    "plt.legend()\n",
    "plt.show()\n",
    "\n",
    "#print(\"history\",history.history)\n",
    "plt.plot(history.history['sparse_categorical_accuracy'],label=\"train\")\n",
    "plt.plot(history.history['val_sparse_categorical_accuracy'],label=\"validation\")\n",
    "plt.title('Model Accuracy')\n",
    "#plt.yscale('log')\n",
    "plt.ylabel('Acc')\n",
    "plt.xlabel('Iteration')\n",
    "plt.grid()\n",
    "plt.legend()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 4.\n",
    "Pamėginti MNIST duomenims gauti kuo didesnį tikslumą."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# train mnist\n",
    "\n",
    "mnist = tf.keras.datasets.mnist\n",
    "(train_images0, train_labels0), (test_images0, test_labels0) = mnist.load_data()\n",
    "\n",
    "test_images  = test_images0.reshape(10000, 784)\n",
    "train_images = train_images0.reshape(60000, 784)\n",
    "\n",
    "test_images  = test_images/255.0\n",
    "train_images = train_images/255.0\n",
    "\n",
    "keras_model = tf.keras.models.Sequential([\n",
    "  tf.keras.layers.Dense(10, activation='softmax')\n",
    "])\n",
    "\n",
    "keras_model.build(input_shape=[None,784])\n",
    "\n",
    "keras_model.summary()\n",
    "\n",
    "keras_model.compile(\n",
    "    optimizer=tf.keras.optimizers.SGD(0.4, use_ema=True),\n",
    "    loss=tf.keras.losses.SparseCategoricalCrossentropy(),\n",
    "    metrics=[tf.keras.metrics.SparseCategoricalAccuracy()],\n",
    ")\n",
    "\n",
    "# Train loop\n",
    "history = keras_model.fit(\n",
    "    train_images,\n",
    "    train_labels0,\n",
    "    batch_size=len(train_images),\n",
    "    epochs=300,\n",
    "    validation_data=(test_images, test_labels0),\n",
    ")\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 60,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Key: loss\n",
      "Key: sparse_categorical_accuracy\n",
      "Key: val_loss\n",
      "Key: val_sparse_categorical_accuracy\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# vizualize mnist\n",
    "\n",
    "for item in history.history:\n",
    "    print(\"Key:\",item)\n",
    "\n",
    "plt.plot(history.history['loss'],label=\"train\")\n",
    "plt.plot(history.history['val_loss'],label=\"validation\")\n",
    "plt.title('Model Loss')\n",
    "plt.yscale('log')\n",
    "plt.ylabel('Cross Entropy')\n",
    "plt.xlabel('Iteration')\n",
    "plt.grid()\n",
    "plt.legend()\n",
    "plt.show()\n",
    "\n",
    "#print(\"history\",history.history)\n",
    "plt.plot(history.history['sparse_categorical_accuracy'],label=\"train\")\n",
    "plt.plot(history.history['val_sparse_categorical_accuracy'],label=\"validation\")\n",
    "plt.title('Model Accuracy')\n",
    "#plt.yscale('log')\n",
    "plt.ylabel('Acc')\n",
    "plt.xlabel('Iteration')\n",
    "plt.grid()\n",
    "plt.legend()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 5.\n",
    "Rasti internete duomenų rinkinį su vaizdais ir pritaikyti turimą kodą."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 81,
   "metadata": {},
   "outputs": [],
   "source": [
    "# load datasets\n",
    "\n",
    "# https://www.kaggle.com/datasets/zalando-research/fashionmnist\n",
    "train_dataset = pd.read_csv('assets/fashion-mnist_train.csv', skiprows = [0], header=None).values\n",
    "train_input = train_dataset[:, 1:]\n",
    "train_label = train_dataset[:, :1]\n",
    "\n",
    "test_dataset = pd.read_csv('assets/fashion-mnist_test.csv', skiprows = [0], header=None).values\n",
    "test_input = test_dataset[:, 1:]\n",
    "test_label = test_dataset[:, :1]\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 88,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model: \"sequential_37\"\n",
      "_________________________________________________________________\n",
      " Layer (type)                Output Shape              Param #   \n",
      "=================================================================\n",
      " dense_56 (Dense)            (None, 10)                7850      \n",
      "                                                                 \n",
      "=================================================================\n",
      "Total params: 7850 (30.66 KB)\n",
      "Trainable params: 7850 (30.66 KB)\n",
      "Non-trainable params: 0 (0.00 Byte)\n",
      "_________________________________________________________________\n",
      "Epoch 1/300\n",
      "1/1 [==============================] - 11s 11s/step - loss: 200.8899 - sparse_categorical_accuracy: 0.0993 - val_loss: 135.2529 - val_sparse_categorical_accuracy: 0.1143\n",
      "Epoch 2/300\n",
      "1/1 [==============================] - 0s 393ms/step - loss: 136.3396 - sparse_categorical_accuracy: 0.1098 - val_loss: 113.3213 - val_sparse_categorical_accuracy: 0.1422\n",
      "Epoch 3/300\n",
      "1/1 [==============================] - 0s 212ms/step - loss: 114.0896 - sparse_categorical_accuracy: 0.1405 - val_loss: 104.0747 - val_sparse_categorical_accuracy: 0.1805\n",
      "Epoch 4/300\n",
      "1/1 [==============================] - 0s 206ms/step - loss: 104.5925 - sparse_categorical_accuracy: 0.1804 - val_loss: 95.5190 - val_sparse_categorical_accuracy: 0.2200\n",
      "Epoch 5/300\n",
      "1/1 [==============================] - 0s 190ms/step - loss: 95.9215 - sparse_categorical_accuracy: 0.2210 - val_loss: 87.2152 - val_sparse_categorical_accuracy: 0.2655\n",
      "Epoch 6/300\n",
      "1/1 [==============================] - 0s 187ms/step - loss: 87.7617 - sparse_categorical_accuracy: 0.2698 - val_loss: 77.3627 - val_sparse_categorical_accuracy: 0.3068\n",
      "Epoch 7/300\n",
      "1/1 [==============================] - 0s 167ms/step - loss: 78.1472 - sparse_categorical_accuracy: 0.3109 - val_loss: 66.3837 - val_sparse_categorical_accuracy: 0.3583\n",
      "Epoch 8/300\n",
      "1/1 [==============================] - 0s 216ms/step - loss: 67.2765 - sparse_categorical_accuracy: 0.3551 - val_loss: 56.1239 - val_sparse_categorical_accuracy: 0.4027\n",
      "Epoch 9/300\n",
      "1/1 [==============================] - 0s 236ms/step - loss: 57.0718 - sparse_categorical_accuracy: 0.4009 - val_loss: 48.2395 - val_sparse_categorical_accuracy: 0.4350\n",
      "Epoch 10/300\n",
      "1/1 [==============================] - 0s 198ms/step - loss: 49.2051 - sparse_categorical_accuracy: 0.4325 - val_loss: 44.2210 - val_sparse_categorical_accuracy: 0.4495\n",
      "Epoch 11/300\n",
      "1/1 [==============================] - 0s 278ms/step - loss: 45.0963 - sparse_categorical_accuracy: 0.4487 - val_loss: 43.6507 - val_sparse_categorical_accuracy: 0.4584\n",
      "Epoch 12/300\n",
      "1/1 [==============================] - 0s 251ms/step - loss: 44.3950 - sparse_categorical_accuracy: 0.4548 - val_loss: 43.7409 - val_sparse_categorical_accuracy: 0.4641\n",
      "Epoch 13/300\n",
      "1/1 [==============================] - 0s 467ms/step - loss: 44.4352 - sparse_categorical_accuracy: 0.4608 - val_loss: 41.6169 - val_sparse_categorical_accuracy: 0.4838\n",
      "Epoch 14/300\n",
      "1/1 [==============================] - 0s 224ms/step - loss: 42.2697 - sparse_categorical_accuracy: 0.4777 - val_loss: 37.5434 - val_sparse_categorical_accuracy: 0.5103\n",
      "Epoch 15/300\n",
      "1/1 [==============================] - 0s 313ms/step - loss: 38.2246 - sparse_categorical_accuracy: 0.5052 - val_loss: 34.0593 - val_sparse_categorical_accuracy: 0.5321\n",
      "Epoch 16/300\n",
      "1/1 [==============================] - 0s 329ms/step - loss: 34.7490 - sparse_categorical_accuracy: 0.5330 - val_loss: 32.4519 - val_sparse_categorical_accuracy: 0.5513\n",
      "Epoch 17/300\n",
      "1/1 [==============================] - 0s 309ms/step - loss: 33.1500 - sparse_categorical_accuracy: 0.5514 - val_loss: 31.9879 - val_sparse_categorical_accuracy: 0.5625\n",
      "Epoch 18/300\n",
      "1/1 [==============================] - 0s 319ms/step - loss: 32.7047 - sparse_categorical_accuracy: 0.5631 - val_loss: 31.5895 - val_sparse_categorical_accuracy: 0.5743\n",
      "Epoch 19/300\n",
      "1/1 [==============================] - 0s 261ms/step - loss: 32.2759 - sparse_categorical_accuracy: 0.5731 - val_loss: 30.6852 - val_sparse_categorical_accuracy: 0.5836\n",
      "Epoch 20/300\n",
      "1/1 [==============================] - 0s 356ms/step - loss: 31.3555 - sparse_categorical_accuracy: 0.5829 - val_loss: 29.3388 - val_sparse_categorical_accuracy: 0.5963\n",
      "Epoch 21/300\n",
      "1/1 [==============================] - 0s 294ms/step - loss: 29.9866 - sparse_categorical_accuracy: 0.5944 - val_loss: 27.9010 - val_sparse_categorical_accuracy: 0.6068\n",
      "Epoch 22/300\n",
      "1/1 [==============================] - 0s 290ms/step - loss: 28.5080 - sparse_categorical_accuracy: 0.6052 - val_loss: 26.8371 - val_sparse_categorical_accuracy: 0.6170\n",
      "Epoch 23/300\n",
      "1/1 [==============================] - 0s 292ms/step - loss: 27.4188 - sparse_categorical_accuracy: 0.6167 - val_loss: 26.3897 - val_sparse_categorical_accuracy: 0.6260\n",
      "Epoch 24/300\n",
      "1/1 [==============================] - 0s 288ms/step - loss: 26.9111 - sparse_categorical_accuracy: 0.6238 - val_loss: 26.0771 - val_sparse_categorical_accuracy: 0.6314\n",
      "Epoch 25/300\n",
      "1/1 [==============================] - 0s 269ms/step - loss: 26.5634 - sparse_categorical_accuracy: 0.6270 - val_loss: 25.4394 - val_sparse_categorical_accuracy: 0.6357\n",
      "Epoch 26/300\n",
      "1/1 [==============================] - 0s 261ms/step - loss: 25.9387 - sparse_categorical_accuracy: 0.6301 - val_loss: 24.5883 - val_sparse_categorical_accuracy: 0.6359\n",
      "Epoch 27/300\n",
      "1/1 [==============================] - 0s 246ms/step - loss: 25.1048 - sparse_categorical_accuracy: 0.6338 - val_loss: 23.8745 - val_sparse_categorical_accuracy: 0.6366\n",
      "Epoch 28/300\n",
      "1/1 [==============================] - 0s 272ms/step - loss: 24.4025 - sparse_categorical_accuracy: 0.6360 - val_loss: 23.4751 - val_sparse_categorical_accuracy: 0.6351\n",
      "Epoch 29/300\n",
      "1/1 [==============================] - 0s 241ms/step - loss: 23.9629 - sparse_categorical_accuracy: 0.6384 - val_loss: 23.1136 - val_sparse_categorical_accuracy: 0.6405\n",
      "Epoch 30/300\n",
      "1/1 [==============================] - 0s 252ms/step - loss: 23.5747 - sparse_categorical_accuracy: 0.6417 - val_loss: 22.6080 - val_sparse_categorical_accuracy: 0.6459\n",
      "Epoch 31/300\n",
      "1/1 [==============================] - 0s 251ms/step - loss: 23.0632 - sparse_categorical_accuracy: 0.6481 - val_loss: 22.0341 - val_sparse_categorical_accuracy: 0.6530\n",
      "Epoch 32/300\n",
      "1/1 [==============================] - 0s 266ms/step - loss: 22.5126 - sparse_categorical_accuracy: 0.6557 - val_loss: 21.6302 - val_sparse_categorical_accuracy: 0.6590\n",
      "Epoch 33/300\n",
      "1/1 [==============================] - 0s 233ms/step - loss: 22.1279 - sparse_categorical_accuracy: 0.6616 - val_loss: 21.4182 - val_sparse_categorical_accuracy: 0.6655\n",
      "Epoch 34/300\n",
      "1/1 [==============================] - 0s 252ms/step - loss: 21.9163 - sparse_categorical_accuracy: 0.6659 - val_loss: 21.1914 - val_sparse_categorical_accuracy: 0.6709\n",
      "Epoch 35/300\n",
      "1/1 [==============================] - 0s 235ms/step - loss: 21.6713 - sparse_categorical_accuracy: 0.6700 - val_loss: 20.7828 - val_sparse_categorical_accuracy: 0.6756\n",
      "Epoch 36/300\n",
      "1/1 [==============================] - 0s 215ms/step - loss: 21.2508 - sparse_categorical_accuracy: 0.6733 - val_loss: 20.2652 - val_sparse_categorical_accuracy: 0.6813\n",
      "Epoch 37/300\n",
      "1/1 [==============================] - 0s 242ms/step - loss: 20.7232 - sparse_categorical_accuracy: 0.6778 - val_loss: 19.8116 - val_sparse_categorical_accuracy: 0.6835\n",
      "Epoch 38/300\n",
      "1/1 [==============================] - 0s 225ms/step - loss: 20.2661 - sparse_categorical_accuracy: 0.6806 - val_loss: 19.5268 - val_sparse_categorical_accuracy: 0.6848\n",
      "Epoch 39/300\n",
      "1/1 [==============================] - 0s 236ms/step - loss: 19.9708 - sparse_categorical_accuracy: 0.6821 - val_loss: 19.3480 - val_sparse_categorical_accuracy: 0.6851\n",
      "Epoch 40/300\n",
      "1/1 [==============================] - 0s 254ms/step - loss: 19.7887 - sparse_categorical_accuracy: 0.6814 - val_loss: 19.1417 - val_sparse_categorical_accuracy: 0.6858\n",
      "Epoch 41/300\n",
      "1/1 [==============================] - 0s 232ms/step - loss: 19.5808 - sparse_categorical_accuracy: 0.6820 - val_loss: 18.8611 - val_sparse_categorical_accuracy: 0.6890\n",
      "Epoch 42/300\n",
      "1/1 [==============================] - 0s 230ms/step - loss: 19.2956 - sparse_categorical_accuracy: 0.6829 - val_loss: 18.5724 - val_sparse_categorical_accuracy: 0.6895\n",
      "Epoch 43/300\n",
      "1/1 [==============================] - 0s 224ms/step - loss: 18.9996 - sparse_categorical_accuracy: 0.6857 - val_loss: 18.3326 - val_sparse_categorical_accuracy: 0.6923\n",
      "Epoch 44/300\n",
      "1/1 [==============================] - 0s 234ms/step - loss: 18.7523 - sparse_categorical_accuracy: 0.6887 - val_loss: 18.1063 - val_sparse_categorical_accuracy: 0.6952\n",
      "Epoch 45/300\n",
      "1/1 [==============================] - 0s 231ms/step - loss: 18.5292 - sparse_categorical_accuracy: 0.6921 - val_loss: 17.8600 - val_sparse_categorical_accuracy: 0.6983\n",
      "Epoch 46/300\n",
      "1/1 [==============================] - 0s 256ms/step - loss: 18.2929 - sparse_categorical_accuracy: 0.6942 - val_loss: 17.6157 - val_sparse_categorical_accuracy: 0.7013\n",
      "Epoch 47/300\n",
      "1/1 [==============================] - 0s 241ms/step - loss: 18.0623 - sparse_categorical_accuracy: 0.6962 - val_loss: 17.4156 - val_sparse_categorical_accuracy: 0.7024\n",
      "Epoch 48/300\n",
      "1/1 [==============================] - 0s 247ms/step - loss: 17.8715 - sparse_categorical_accuracy: 0.6980 - val_loss: 17.2511 - val_sparse_categorical_accuracy: 0.7028\n",
      "Epoch 49/300\n",
      "1/1 [==============================] - 0s 232ms/step - loss: 17.7106 - sparse_categorical_accuracy: 0.6991 - val_loss: 17.0785 - val_sparse_categorical_accuracy: 0.7055\n",
      "Epoch 50/300\n",
      "1/1 [==============================] - 0s 245ms/step - loss: 17.5329 - sparse_categorical_accuracy: 0.7005 - val_loss: 16.8799 - val_sparse_categorical_accuracy: 0.7073\n",
      "Epoch 51/300\n",
      "1/1 [==============================] - 0s 237ms/step - loss: 17.3180 - sparse_categorical_accuracy: 0.7023 - val_loss: 16.6840 - val_sparse_categorical_accuracy: 0.7105\n",
      "Epoch 52/300\n",
      "1/1 [==============================] - 0s 230ms/step - loss: 17.0982 - sparse_categorical_accuracy: 0.7030 - val_loss: 16.5309 - val_sparse_categorical_accuracy: 0.7110\n",
      "Epoch 53/300\n",
      "1/1 [==============================] - 0s 230ms/step - loss: 16.9205 - sparse_categorical_accuracy: 0.7032 - val_loss: 16.4061 - val_sparse_categorical_accuracy: 0.7117\n",
      "Epoch 54/300\n",
      "1/1 [==============================] - 0s 231ms/step - loss: 16.7789 - sparse_categorical_accuracy: 0.7033 - val_loss: 16.2694 - val_sparse_categorical_accuracy: 0.7112\n",
      "Epoch 55/300\n",
      "1/1 [==============================] - 0s 240ms/step - loss: 16.6307 - sparse_categorical_accuracy: 0.7042 - val_loss: 16.1078 - val_sparse_categorical_accuracy: 0.7120\n",
      "Epoch 56/300\n",
      "1/1 [==============================] - 0s 244ms/step - loss: 16.4600 - sparse_categorical_accuracy: 0.7050 - val_loss: 15.9454 - val_sparse_categorical_accuracy: 0.7137\n",
      "Epoch 57/300\n",
      "1/1 [==============================] - 0s 231ms/step - loss: 16.2947 - sparse_categorical_accuracy: 0.7066 - val_loss: 15.8032 - val_sparse_categorical_accuracy: 0.7137\n",
      "Epoch 58/300\n",
      "1/1 [==============================] - 0s 229ms/step - loss: 16.1502 - sparse_categorical_accuracy: 0.7081 - val_loss: 15.6730 - val_sparse_categorical_accuracy: 0.7143\n",
      "Epoch 59/300\n",
      "1/1 [==============================] - 0s 236ms/step - loss: 16.0130 - sparse_categorical_accuracy: 0.7093 - val_loss: 15.5380 - val_sparse_categorical_accuracy: 0.7161\n",
      "Epoch 60/300\n",
      "1/1 [==============================] - 0s 215ms/step - loss: 15.8696 - sparse_categorical_accuracy: 0.7104 - val_loss: 15.4090 - val_sparse_categorical_accuracy: 0.7182\n",
      "Epoch 61/300\n",
      "1/1 [==============================] - 0s 233ms/step - loss: 15.7291 - sparse_categorical_accuracy: 0.7122 - val_loss: 15.2951 - val_sparse_categorical_accuracy: 0.7203\n",
      "Epoch 62/300\n",
      "1/1 [==============================] - 0s 222ms/step - loss: 15.6048 - sparse_categorical_accuracy: 0.7139 - val_loss: 15.1881 - val_sparse_categorical_accuracy: 0.7224\n",
      "Epoch 63/300\n",
      "1/1 [==============================] - 0s 238ms/step - loss: 15.4892 - sparse_categorical_accuracy: 0.7150 - val_loss: 15.0698 - val_sparse_categorical_accuracy: 0.7237\n",
      "Epoch 64/300\n",
      "1/1 [==============================] - 0s 248ms/step - loss: 15.3673 - sparse_categorical_accuracy: 0.7162 - val_loss: 14.9416 - val_sparse_categorical_accuracy: 0.7250\n",
      "Epoch 65/300\n",
      "1/1 [==============================] - 0s 266ms/step - loss: 15.2410 - sparse_categorical_accuracy: 0.7167 - val_loss: 14.8211 - val_sparse_categorical_accuracy: 0.7247\n",
      "Epoch 66/300\n",
      "1/1 [==============================] - 0s 275ms/step - loss: 15.1234 - sparse_categorical_accuracy: 0.7171 - val_loss: 14.7128 - val_sparse_categorical_accuracy: 0.7266\n",
      "Epoch 67/300\n",
      "1/1 [==============================] - 0s 246ms/step - loss: 15.0171 - sparse_categorical_accuracy: 0.7182 - val_loss: 14.6074 - val_sparse_categorical_accuracy: 0.7275\n",
      "Epoch 68/300\n",
      "1/1 [==============================] - 0s 250ms/step - loss: 14.9121 - sparse_categorical_accuracy: 0.7195 - val_loss: 14.4997 - val_sparse_categorical_accuracy: 0.7297\n",
      "Epoch 69/300\n",
      "1/1 [==============================] - 0s 243ms/step - loss: 14.8028 - sparse_categorical_accuracy: 0.7206 - val_loss: 14.3956 - val_sparse_categorical_accuracy: 0.7313\n",
      "Epoch 70/300\n",
      "1/1 [==============================] - 0s 246ms/step - loss: 14.6953 - sparse_categorical_accuracy: 0.7217 - val_loss: 14.2986 - val_sparse_categorical_accuracy: 0.7324\n",
      "Epoch 71/300\n",
      "1/1 [==============================] - 0s 338ms/step - loss: 14.5937 - sparse_categorical_accuracy: 0.7230 - val_loss: 14.2029 - val_sparse_categorical_accuracy: 0.7323\n",
      "Epoch 72/300\n",
      "1/1 [==============================] - 0s 283ms/step - loss: 14.4943 - sparse_categorical_accuracy: 0.7243 - val_loss: 14.1052 - val_sparse_categorical_accuracy: 0.7333\n",
      "Epoch 73/300\n",
      "1/1 [==============================] - 0s 261ms/step - loss: 14.3949 - sparse_categorical_accuracy: 0.7255 - val_loss: 14.0099 - val_sparse_categorical_accuracy: 0.7336\n",
      "Epoch 74/300\n",
      "1/1 [==============================] - 0s 262ms/step - loss: 14.2991 - sparse_categorical_accuracy: 0.7263 - val_loss: 13.9184 - val_sparse_categorical_accuracy: 0.7351\n",
      "Epoch 75/300\n",
      "1/1 [==============================] - 0s 273ms/step - loss: 14.2084 - sparse_categorical_accuracy: 0.7267 - val_loss: 13.8273 - val_sparse_categorical_accuracy: 0.7352\n",
      "Epoch 76/300\n",
      "1/1 [==============================] - 0s 261ms/step - loss: 14.1187 - sparse_categorical_accuracy: 0.7277 - val_loss: 13.7361 - val_sparse_categorical_accuracy: 0.7365\n",
      "Epoch 77/300\n",
      "1/1 [==============================] - 0s 269ms/step - loss: 14.0274 - sparse_categorical_accuracy: 0.7285 - val_loss: 13.6494 - val_sparse_categorical_accuracy: 0.7375\n",
      "Epoch 78/300\n",
      "1/1 [==============================] - 0s 282ms/step - loss: 13.9381 - sparse_categorical_accuracy: 0.7294 - val_loss: 13.5701 - val_sparse_categorical_accuracy: 0.7388\n",
      "Epoch 79/300\n",
      "1/1 [==============================] - 0s 308ms/step - loss: 13.8536 - sparse_categorical_accuracy: 0.7304 - val_loss: 13.4938 - val_sparse_categorical_accuracy: 0.7389\n",
      "Epoch 80/300\n",
      "1/1 [==============================] - 0s 258ms/step - loss: 13.7718 - sparse_categorical_accuracy: 0.7315 - val_loss: 13.4165 - val_sparse_categorical_accuracy: 0.7390\n",
      "Epoch 81/300\n",
      "1/1 [==============================] - 0s 261ms/step - loss: 13.6900 - sparse_categorical_accuracy: 0.7324 - val_loss: 13.3395 - val_sparse_categorical_accuracy: 0.7407\n",
      "Epoch 82/300\n",
      "1/1 [==============================] - 0s 266ms/step - loss: 13.6093 - sparse_categorical_accuracy: 0.7333 - val_loss: 13.2646 - val_sparse_categorical_accuracy: 0.7416\n",
      "Epoch 83/300\n",
      "1/1 [==============================] - 0s 284ms/step - loss: 13.5312 - sparse_categorical_accuracy: 0.7342 - val_loss: 13.1913 - val_sparse_categorical_accuracy: 0.7418\n",
      "Epoch 84/300\n",
      "1/1 [==============================] - 0s 250ms/step - loss: 13.4549 - sparse_categorical_accuracy: 0.7351 - val_loss: 13.1195 - val_sparse_categorical_accuracy: 0.7427\n",
      "Epoch 85/300\n",
      "1/1 [==============================] - 0s 252ms/step - loss: 13.3794 - sparse_categorical_accuracy: 0.7357 - val_loss: 13.0506 - val_sparse_categorical_accuracy: 0.7437\n",
      "Epoch 86/300\n",
      "1/1 [==============================] - 0s 263ms/step - loss: 13.3061 - sparse_categorical_accuracy: 0.7363 - val_loss: 12.9844 - val_sparse_categorical_accuracy: 0.7443\n",
      "Epoch 87/300\n",
      "1/1 [==============================] - 0s 251ms/step - loss: 13.2355 - sparse_categorical_accuracy: 0.7367 - val_loss: 12.9177 - val_sparse_categorical_accuracy: 0.7446\n",
      "Epoch 88/300\n",
      "1/1 [==============================] - 0s 327ms/step - loss: 13.1651 - sparse_categorical_accuracy: 0.7371 - val_loss: 12.8496 - val_sparse_categorical_accuracy: 0.7448\n",
      "Epoch 89/300\n",
      "1/1 [==============================] - 0s 252ms/step - loss: 13.0944 - sparse_categorical_accuracy: 0.7376 - val_loss: 12.7826 - val_sparse_categorical_accuracy: 0.7442\n",
      "Epoch 90/300\n",
      "1/1 [==============================] - 0s 252ms/step - loss: 13.0254 - sparse_categorical_accuracy: 0.7385 - val_loss: 12.7189 - val_sparse_categorical_accuracy: 0.7451\n",
      "Epoch 91/300\n",
      "1/1 [==============================] - 0s 261ms/step - loss: 12.9589 - sparse_categorical_accuracy: 0.7395 - val_loss: 12.6576 - val_sparse_categorical_accuracy: 0.7457\n",
      "Epoch 92/300\n",
      "1/1 [==============================] - 0s 253ms/step - loss: 12.8934 - sparse_categorical_accuracy: 0.7402 - val_loss: 12.5982 - val_sparse_categorical_accuracy: 0.7464\n",
      "Epoch 93/300\n",
      "1/1 [==============================] - 0s 240ms/step - loss: 12.8282 - sparse_categorical_accuracy: 0.7410 - val_loss: 12.5400 - val_sparse_categorical_accuracy: 0.7470\n",
      "Epoch 94/300\n",
      "1/1 [==============================] - 0s 251ms/step - loss: 12.7640 - sparse_categorical_accuracy: 0.7416 - val_loss: 12.4820 - val_sparse_categorical_accuracy: 0.7474\n",
      "Epoch 95/300\n",
      "1/1 [==============================] - 0s 262ms/step - loss: 12.7009 - sparse_categorical_accuracy: 0.7422 - val_loss: 12.4235 - val_sparse_categorical_accuracy: 0.7474\n",
      "Epoch 96/300\n",
      "1/1 [==============================] - 0s 244ms/step - loss: 12.6387 - sparse_categorical_accuracy: 0.7426 - val_loss: 12.3656 - val_sparse_categorical_accuracy: 0.7477\n",
      "Epoch 97/300\n",
      "1/1 [==============================] - 0s 269ms/step - loss: 12.5777 - sparse_categorical_accuracy: 0.7426 - val_loss: 12.3098 - val_sparse_categorical_accuracy: 0.7485\n",
      "Epoch 98/300\n",
      "1/1 [==============================] - 0s 247ms/step - loss: 12.5179 - sparse_categorical_accuracy: 0.7429 - val_loss: 12.2564 - val_sparse_categorical_accuracy: 0.7493\n",
      "Epoch 99/300\n",
      "1/1 [==============================] - 0s 257ms/step - loss: 12.4585 - sparse_categorical_accuracy: 0.7434 - val_loss: 12.2053 - val_sparse_categorical_accuracy: 0.7497\n",
      "Epoch 100/300\n",
      "1/1 [==============================] - 0s 243ms/step - loss: 12.3994 - sparse_categorical_accuracy: 0.7443 - val_loss: 12.1561 - val_sparse_categorical_accuracy: 0.7500\n",
      "Epoch 101/300\n",
      "1/1 [==============================] - 0s 247ms/step - loss: 12.3415 - sparse_categorical_accuracy: 0.7448 - val_loss: 12.1070 - val_sparse_categorical_accuracy: 0.7502\n",
      "Epoch 102/300\n",
      "1/1 [==============================] - 0s 262ms/step - loss: 12.2845 - sparse_categorical_accuracy: 0.7454 - val_loss: 12.0567 - val_sparse_categorical_accuracy: 0.7508\n",
      "Epoch 103/300\n",
      "1/1 [==============================] - 0s 258ms/step - loss: 12.2280 - sparse_categorical_accuracy: 0.7460 - val_loss: 12.0055 - val_sparse_categorical_accuracy: 0.7510\n",
      "Epoch 104/300\n",
      "1/1 [==============================] - 0s 255ms/step - loss: 12.1720 - sparse_categorical_accuracy: 0.7463 - val_loss: 11.9552 - val_sparse_categorical_accuracy: 0.7511\n",
      "Epoch 105/300\n",
      "1/1 [==============================] - 0s 249ms/step - loss: 12.1167 - sparse_categorical_accuracy: 0.7468 - val_loss: 11.9066 - val_sparse_categorical_accuracy: 0.7515\n",
      "Epoch 106/300\n",
      "1/1 [==============================] - 0s 256ms/step - loss: 12.0620 - sparse_categorical_accuracy: 0.7471 - val_loss: 11.8599 - val_sparse_categorical_accuracy: 0.7517\n",
      "Epoch 107/300\n",
      "1/1 [==============================] - 0s 286ms/step - loss: 12.0079 - sparse_categorical_accuracy: 0.7478 - val_loss: 11.8144 - val_sparse_categorical_accuracy: 0.7525\n",
      "Epoch 108/300\n",
      "1/1 [==============================] - 0s 255ms/step - loss: 11.9545 - sparse_categorical_accuracy: 0.7481 - val_loss: 11.7689 - val_sparse_categorical_accuracy: 0.7527\n",
      "Epoch 109/300\n",
      "1/1 [==============================] - 0s 275ms/step - loss: 11.9018 - sparse_categorical_accuracy: 0.7485 - val_loss: 11.7227 - val_sparse_categorical_accuracy: 0.7532\n",
      "Epoch 110/300\n",
      "1/1 [==============================] - 0s 245ms/step - loss: 11.8496 - sparse_categorical_accuracy: 0.7491 - val_loss: 11.6758 - val_sparse_categorical_accuracy: 0.7537\n",
      "Epoch 111/300\n",
      "1/1 [==============================] - 0s 268ms/step - loss: 11.7978 - sparse_categorical_accuracy: 0.7498 - val_loss: 11.6294 - val_sparse_categorical_accuracy: 0.7534\n",
      "Epoch 112/300\n",
      "1/1 [==============================] - 0s 241ms/step - loss: 11.7467 - sparse_categorical_accuracy: 0.7502 - val_loss: 11.5841 - val_sparse_categorical_accuracy: 0.7540\n",
      "Epoch 113/300\n",
      "1/1 [==============================] - 0s 262ms/step - loss: 11.6962 - sparse_categorical_accuracy: 0.7506 - val_loss: 11.5399 - val_sparse_categorical_accuracy: 0.7543\n",
      "Epoch 114/300\n",
      "1/1 [==============================] - 0s 265ms/step - loss: 11.6461 - sparse_categorical_accuracy: 0.7510 - val_loss: 11.4964 - val_sparse_categorical_accuracy: 0.7550\n",
      "Epoch 115/300\n",
      "1/1 [==============================] - 0s 328ms/step - loss: 11.5966 - sparse_categorical_accuracy: 0.7515 - val_loss: 11.4529 - val_sparse_categorical_accuracy: 0.7552\n",
      "Epoch 116/300\n",
      "1/1 [==============================] - 0s 272ms/step - loss: 11.5475 - sparse_categorical_accuracy: 0.7519 - val_loss: 11.4092 - val_sparse_categorical_accuracy: 0.7558\n",
      "Epoch 117/300\n",
      "1/1 [==============================] - 0s 289ms/step - loss: 11.4990 - sparse_categorical_accuracy: 0.7523 - val_loss: 11.3654 - val_sparse_categorical_accuracy: 0.7559\n",
      "Epoch 118/300\n",
      "1/1 [==============================] - 0s 280ms/step - loss: 11.4509 - sparse_categorical_accuracy: 0.7526 - val_loss: 11.3225 - val_sparse_categorical_accuracy: 0.7565\n",
      "Epoch 119/300\n",
      "1/1 [==============================] - 0s 257ms/step - loss: 11.4034 - sparse_categorical_accuracy: 0.7528 - val_loss: 11.2807 - val_sparse_categorical_accuracy: 0.7564\n",
      "Epoch 120/300\n",
      "1/1 [==============================] - 0s 253ms/step - loss: 11.3562 - sparse_categorical_accuracy: 0.7533 - val_loss: 11.2400 - val_sparse_categorical_accuracy: 0.7567\n",
      "Epoch 121/300\n",
      "1/1 [==============================] - 0s 251ms/step - loss: 11.3095 - sparse_categorical_accuracy: 0.7539 - val_loss: 11.2000 - val_sparse_categorical_accuracy: 0.7566\n",
      "Epoch 122/300\n",
      "1/1 [==============================] - 0s 259ms/step - loss: 11.2633 - sparse_categorical_accuracy: 0.7543 - val_loss: 11.1602 - val_sparse_categorical_accuracy: 0.7568\n",
      "Epoch 123/300\n",
      "1/1 [==============================] - 0s 258ms/step - loss: 11.2176 - sparse_categorical_accuracy: 0.7547 - val_loss: 11.1202 - val_sparse_categorical_accuracy: 0.7571\n",
      "Epoch 124/300\n",
      "1/1 [==============================] - 0s 263ms/step - loss: 11.1722 - sparse_categorical_accuracy: 0.7551 - val_loss: 11.0806 - val_sparse_categorical_accuracy: 0.7574\n",
      "Epoch 125/300\n",
      "1/1 [==============================] - 0s 256ms/step - loss: 11.1274 - sparse_categorical_accuracy: 0.7555 - val_loss: 11.0417 - val_sparse_categorical_accuracy: 0.7581\n",
      "Epoch 126/300\n",
      "1/1 [==============================] - 0s 272ms/step - loss: 11.0830 - sparse_categorical_accuracy: 0.7559 - val_loss: 11.0037 - val_sparse_categorical_accuracy: 0.7584\n",
      "Epoch 127/300\n",
      "1/1 [==============================] - 0s 253ms/step - loss: 11.0391 - sparse_categorical_accuracy: 0.7563 - val_loss: 10.9663 - val_sparse_categorical_accuracy: 0.7589\n",
      "Epoch 128/300\n",
      "1/1 [==============================] - 0s 274ms/step - loss: 10.9955 - sparse_categorical_accuracy: 0.7564 - val_loss: 10.9290 - val_sparse_categorical_accuracy: 0.7595\n",
      "Epoch 129/300\n",
      "1/1 [==============================] - 0s 276ms/step - loss: 10.9525 - sparse_categorical_accuracy: 0.7570 - val_loss: 10.8914 - val_sparse_categorical_accuracy: 0.7598\n",
      "Epoch 130/300\n",
      "1/1 [==============================] - 0s 263ms/step - loss: 10.9098 - sparse_categorical_accuracy: 0.7573 - val_loss: 10.8537 - val_sparse_categorical_accuracy: 0.7601\n",
      "Epoch 131/300\n",
      "1/1 [==============================] - 0s 250ms/step - loss: 10.8675 - sparse_categorical_accuracy: 0.7578 - val_loss: 10.8164 - val_sparse_categorical_accuracy: 0.7600\n",
      "Epoch 132/300\n",
      "1/1 [==============================] - 0s 275ms/step - loss: 10.8257 - sparse_categorical_accuracy: 0.7582 - val_loss: 10.7798 - val_sparse_categorical_accuracy: 0.7600\n",
      "Epoch 133/300\n",
      "1/1 [==============================] - 0s 237ms/step - loss: 10.7842 - sparse_categorical_accuracy: 0.7586 - val_loss: 10.7436 - val_sparse_categorical_accuracy: 0.7603\n",
      "Epoch 134/300\n",
      "1/1 [==============================] - 0s 245ms/step - loss: 10.7432 - sparse_categorical_accuracy: 0.7590 - val_loss: 10.7076 - val_sparse_categorical_accuracy: 0.7608\n",
      "Epoch 135/300\n",
      "1/1 [==============================] - 0s 240ms/step - loss: 10.7025 - sparse_categorical_accuracy: 0.7594 - val_loss: 10.6716 - val_sparse_categorical_accuracy: 0.7610\n",
      "Epoch 136/300\n",
      "1/1 [==============================] - 0s 238ms/step - loss: 10.6622 - sparse_categorical_accuracy: 0.7598 - val_loss: 10.6355 - val_sparse_categorical_accuracy: 0.7607\n",
      "Epoch 137/300\n",
      "1/1 [==============================] - 0s 243ms/step - loss: 10.6222 - sparse_categorical_accuracy: 0.7602 - val_loss: 10.5999 - val_sparse_categorical_accuracy: 0.7609\n",
      "Epoch 138/300\n",
      "1/1 [==============================] - 0s 265ms/step - loss: 10.5827 - sparse_categorical_accuracy: 0.7607 - val_loss: 10.5649 - val_sparse_categorical_accuracy: 0.7612\n",
      "Epoch 139/300\n",
      "1/1 [==============================] - 0s 231ms/step - loss: 10.5434 - sparse_categorical_accuracy: 0.7612 - val_loss: 10.5305 - val_sparse_categorical_accuracy: 0.7618\n",
      "Epoch 140/300\n",
      "1/1 [==============================] - 0s 248ms/step - loss: 10.5045 - sparse_categorical_accuracy: 0.7616 - val_loss: 10.4964 - val_sparse_categorical_accuracy: 0.7621\n",
      "Epoch 141/300\n",
      "1/1 [==============================] - 0s 228ms/step - loss: 10.4660 - sparse_categorical_accuracy: 0.7616 - val_loss: 10.4624 - val_sparse_categorical_accuracy: 0.7625\n",
      "Epoch 142/300\n",
      "1/1 [==============================] - 0s 232ms/step - loss: 10.4278 - sparse_categorical_accuracy: 0.7620 - val_loss: 10.4285 - val_sparse_categorical_accuracy: 0.7629\n",
      "Epoch 143/300\n",
      "1/1 [==============================] - 0s 255ms/step - loss: 10.3899 - sparse_categorical_accuracy: 0.7624 - val_loss: 10.3950 - val_sparse_categorical_accuracy: 0.7633\n",
      "Epoch 144/300\n",
      "1/1 [==============================] - 0s 298ms/step - loss: 10.3523 - sparse_categorical_accuracy: 0.7628 - val_loss: 10.3620 - val_sparse_categorical_accuracy: 0.7635\n",
      "Epoch 145/300\n",
      "1/1 [==============================] - 0s 277ms/step - loss: 10.3151 - sparse_categorical_accuracy: 0.7631 - val_loss: 10.3296 - val_sparse_categorical_accuracy: 0.7638\n",
      "Epoch 146/300\n",
      "1/1 [==============================] - 0s 247ms/step - loss: 10.2782 - sparse_categorical_accuracy: 0.7634 - val_loss: 10.2973 - val_sparse_categorical_accuracy: 0.7640\n",
      "Epoch 147/300\n",
      "1/1 [==============================] - 0s 258ms/step - loss: 10.2416 - sparse_categorical_accuracy: 0.7636 - val_loss: 10.2651 - val_sparse_categorical_accuracy: 0.7641\n",
      "Epoch 148/300\n",
      "1/1 [==============================] - 0s 257ms/step - loss: 10.2053 - sparse_categorical_accuracy: 0.7638 - val_loss: 10.2330 - val_sparse_categorical_accuracy: 0.7646\n",
      "Epoch 149/300\n",
      "1/1 [==============================] - 0s 259ms/step - loss: 10.1693 - sparse_categorical_accuracy: 0.7641 - val_loss: 10.2011 - val_sparse_categorical_accuracy: 0.7649\n",
      "Epoch 150/300\n",
      "1/1 [==============================] - 0s 236ms/step - loss: 10.1336 - sparse_categorical_accuracy: 0.7643 - val_loss: 10.1697 - val_sparse_categorical_accuracy: 0.7654\n",
      "Epoch 151/300\n",
      "1/1 [==============================] - 0s 340ms/step - loss: 10.0982 - sparse_categorical_accuracy: 0.7645 - val_loss: 10.1387 - val_sparse_categorical_accuracy: 0.7662\n",
      "Epoch 152/300\n",
      "1/1 [==============================] - 0s 273ms/step - loss: 10.0630 - sparse_categorical_accuracy: 0.7649 - val_loss: 10.1079 - val_sparse_categorical_accuracy: 0.7666\n",
      "Epoch 153/300\n",
      "1/1 [==============================] - 0s 271ms/step - loss: 10.0281 - sparse_categorical_accuracy: 0.7654 - val_loss: 10.0772 - val_sparse_categorical_accuracy: 0.7666\n",
      "Epoch 154/300\n",
      "1/1 [==============================] - 0s 262ms/step - loss: 9.9935 - sparse_categorical_accuracy: 0.7657 - val_loss: 10.0466 - val_sparse_categorical_accuracy: 0.7669\n",
      "Epoch 155/300\n",
      "1/1 [==============================] - 0s 265ms/step - loss: 9.9592 - sparse_categorical_accuracy: 0.7660 - val_loss: 10.0164 - val_sparse_categorical_accuracy: 0.7674\n",
      "Epoch 156/300\n",
      "1/1 [==============================] - 0s 245ms/step - loss: 9.9251 - sparse_categorical_accuracy: 0.7660 - val_loss: 9.9866 - val_sparse_categorical_accuracy: 0.7676\n",
      "Epoch 157/300\n",
      "1/1 [==============================] - 0s 298ms/step - loss: 9.8913 - sparse_categorical_accuracy: 0.7662 - val_loss: 9.9571 - val_sparse_categorical_accuracy: 0.7679\n",
      "Epoch 158/300\n",
      "1/1 [==============================] - 0s 259ms/step - loss: 9.8577 - sparse_categorical_accuracy: 0.7666 - val_loss: 9.9277 - val_sparse_categorical_accuracy: 0.7682\n",
      "Epoch 159/300\n",
      "1/1 [==============================] - 0s 259ms/step - loss: 9.8244 - sparse_categorical_accuracy: 0.7668 - val_loss: 9.8985 - val_sparse_categorical_accuracy: 0.7683\n",
      "Epoch 160/300\n",
      "1/1 [==============================] - 0s 254ms/step - loss: 9.7913 - sparse_categorical_accuracy: 0.7668 - val_loss: 9.8695 - val_sparse_categorical_accuracy: 0.7687\n",
      "Epoch 161/300\n",
      "1/1 [==============================] - 0s 231ms/step - loss: 9.7585 - sparse_categorical_accuracy: 0.7670 - val_loss: 9.8408 - val_sparse_categorical_accuracy: 0.7688\n",
      "Epoch 162/300\n",
      "1/1 [==============================] - 0s 242ms/step - loss: 9.7258 - sparse_categorical_accuracy: 0.7675 - val_loss: 9.8126 - val_sparse_categorical_accuracy: 0.7691\n",
      "Epoch 163/300\n",
      "1/1 [==============================] - 0s 243ms/step - loss: 9.6934 - sparse_categorical_accuracy: 0.7678 - val_loss: 9.7846 - val_sparse_categorical_accuracy: 0.7691\n",
      "Epoch 164/300\n",
      "1/1 [==============================] - 0s 231ms/step - loss: 9.6613 - sparse_categorical_accuracy: 0.7680 - val_loss: 9.7568 - val_sparse_categorical_accuracy: 0.7693\n",
      "Epoch 165/300\n",
      "1/1 [==============================] - 0s 241ms/step - loss: 9.6293 - sparse_categorical_accuracy: 0.7682 - val_loss: 9.7291 - val_sparse_categorical_accuracy: 0.7695\n",
      "Epoch 166/300\n",
      "1/1 [==============================] - 0s 240ms/step - loss: 9.5976 - sparse_categorical_accuracy: 0.7686 - val_loss: 9.7016 - val_sparse_categorical_accuracy: 0.7694\n",
      "Epoch 167/300\n",
      "1/1 [==============================] - 0s 255ms/step - loss: 9.5660 - sparse_categorical_accuracy: 0.7689 - val_loss: 9.6745 - val_sparse_categorical_accuracy: 0.7699\n",
      "Epoch 168/300\n",
      "1/1 [==============================] - 0s 265ms/step - loss: 9.5347 - sparse_categorical_accuracy: 0.7692 - val_loss: 9.6477 - val_sparse_categorical_accuracy: 0.7701\n",
      "Epoch 169/300\n",
      "1/1 [==============================] - 0s 248ms/step - loss: 9.5037 - sparse_categorical_accuracy: 0.7695 - val_loss: 9.6211 - val_sparse_categorical_accuracy: 0.7703\n",
      "Epoch 170/300\n",
      "1/1 [==============================] - 0s 252ms/step - loss: 9.4728 - sparse_categorical_accuracy: 0.7697 - val_loss: 9.5947 - val_sparse_categorical_accuracy: 0.7703\n",
      "Epoch 171/300\n",
      "1/1 [==============================] - 0s 245ms/step - loss: 9.4421 - sparse_categorical_accuracy: 0.7700 - val_loss: 9.5685 - val_sparse_categorical_accuracy: 0.7704\n",
      "Epoch 172/300\n",
      "1/1 [==============================] - 0s 238ms/step - loss: 9.4117 - sparse_categorical_accuracy: 0.7702 - val_loss: 9.5425 - val_sparse_categorical_accuracy: 0.7704\n",
      "Epoch 173/300\n",
      "1/1 [==============================] - 0s 250ms/step - loss: 9.3814 - sparse_categorical_accuracy: 0.7705 - val_loss: 9.5168 - val_sparse_categorical_accuracy: 0.7705\n",
      "Epoch 174/300\n",
      "1/1 [==============================] - 0s 249ms/step - loss: 9.3514 - sparse_categorical_accuracy: 0.7707 - val_loss: 9.4913 - val_sparse_categorical_accuracy: 0.7709\n",
      "Epoch 175/300\n",
      "1/1 [==============================] - 0s 261ms/step - loss: 9.3216 - sparse_categorical_accuracy: 0.7710 - val_loss: 9.4661 - val_sparse_categorical_accuracy: 0.7708\n",
      "Epoch 176/300\n",
      "1/1 [==============================] - 0s 238ms/step - loss: 9.2919 - sparse_categorical_accuracy: 0.7711 - val_loss: 9.4409 - val_sparse_categorical_accuracy: 0.7709\n",
      "Epoch 177/300\n",
      "1/1 [==============================] - 0s 251ms/step - loss: 9.2625 - sparse_categorical_accuracy: 0.7714 - val_loss: 9.4159 - val_sparse_categorical_accuracy: 0.7711\n",
      "Epoch 178/300\n",
      "1/1 [==============================] - 0s 292ms/step - loss: 9.2332 - sparse_categorical_accuracy: 0.7715 - val_loss: 9.3911 - val_sparse_categorical_accuracy: 0.7713\n",
      "Epoch 179/300\n",
      "1/1 [==============================] - 0s 258ms/step - loss: 9.2041 - sparse_categorical_accuracy: 0.7718 - val_loss: 9.3666 - val_sparse_categorical_accuracy: 0.7714\n",
      "Epoch 180/300\n",
      "1/1 [==============================] - 0s 231ms/step - loss: 9.1752 - sparse_categorical_accuracy: 0.7721 - val_loss: 9.3421 - val_sparse_categorical_accuracy: 0.7718\n",
      "Epoch 181/300\n",
      "1/1 [==============================] - 0s 260ms/step - loss: 9.1465 - sparse_categorical_accuracy: 0.7724 - val_loss: 9.3178 - val_sparse_categorical_accuracy: 0.7719\n",
      "Epoch 182/300\n",
      "1/1 [==============================] - 0s 296ms/step - loss: 9.1179 - sparse_categorical_accuracy: 0.7728 - val_loss: 9.2937 - val_sparse_categorical_accuracy: 0.7722\n",
      "Epoch 183/300\n",
      "1/1 [==============================] - 0s 243ms/step - loss: 9.0895 - sparse_categorical_accuracy: 0.7728 - val_loss: 9.2697 - val_sparse_categorical_accuracy: 0.7722\n",
      "Epoch 184/300\n",
      "1/1 [==============================] - 0s 239ms/step - loss: 9.0613 - sparse_categorical_accuracy: 0.7732 - val_loss: 9.2459 - val_sparse_categorical_accuracy: 0.7721\n",
      "Epoch 185/300\n",
      "1/1 [==============================] - 0s 254ms/step - loss: 9.0333 - sparse_categorical_accuracy: 0.7736 - val_loss: 9.2224 - val_sparse_categorical_accuracy: 0.7725\n",
      "Epoch 186/300\n",
      "1/1 [==============================] - 0s 238ms/step - loss: 9.0054 - sparse_categorical_accuracy: 0.7738 - val_loss: 9.1990 - val_sparse_categorical_accuracy: 0.7730\n",
      "Epoch 187/300\n",
      "1/1 [==============================] - 0s 280ms/step - loss: 8.9777 - sparse_categorical_accuracy: 0.7739 - val_loss: 9.1758 - val_sparse_categorical_accuracy: 0.7730\n",
      "Epoch 188/300\n",
      "1/1 [==============================] - 0s 258ms/step - loss: 8.9502 - sparse_categorical_accuracy: 0.7743 - val_loss: 9.1527 - val_sparse_categorical_accuracy: 0.7731\n",
      "Epoch 189/300\n",
      "1/1 [==============================] - 0s 258ms/step - loss: 8.9228 - sparse_categorical_accuracy: 0.7744 - val_loss: 9.1299 - val_sparse_categorical_accuracy: 0.7730\n",
      "Epoch 190/300\n",
      "1/1 [==============================] - 0s 238ms/step - loss: 8.8957 - sparse_categorical_accuracy: 0.7747 - val_loss: 9.1072 - val_sparse_categorical_accuracy: 0.7731\n",
      "Epoch 191/300\n",
      "1/1 [==============================] - 0s 245ms/step - loss: 8.8686 - sparse_categorical_accuracy: 0.7750 - val_loss: 9.0846 - val_sparse_categorical_accuracy: 0.7738\n",
      "Epoch 192/300\n",
      "1/1 [==============================] - 0s 250ms/step - loss: 8.8418 - sparse_categorical_accuracy: 0.7751 - val_loss: 9.0622 - val_sparse_categorical_accuracy: 0.7744\n",
      "Epoch 193/300\n",
      "1/1 [==============================] - 0s 270ms/step - loss: 8.8151 - sparse_categorical_accuracy: 0.7753 - val_loss: 9.0399 - val_sparse_categorical_accuracy: 0.7747\n",
      "Epoch 194/300\n",
      "1/1 [==============================] - 0s 244ms/step - loss: 8.7886 - sparse_categorical_accuracy: 0.7756 - val_loss: 9.0177 - val_sparse_categorical_accuracy: 0.7750\n",
      "Epoch 195/300\n",
      "1/1 [==============================] - 0s 249ms/step - loss: 8.7622 - sparse_categorical_accuracy: 0.7758 - val_loss: 8.9957 - val_sparse_categorical_accuracy: 0.7751\n",
      "Epoch 196/300\n",
      "1/1 [==============================] - 0s 254ms/step - loss: 8.7360 - sparse_categorical_accuracy: 0.7758 - val_loss: 8.9738 - val_sparse_categorical_accuracy: 0.7756\n",
      "Epoch 197/300\n",
      "1/1 [==============================] - 0s 251ms/step - loss: 8.7099 - sparse_categorical_accuracy: 0.7760 - val_loss: 8.9520 - val_sparse_categorical_accuracy: 0.7755\n",
      "Epoch 198/300\n",
      "1/1 [==============================] - 0s 258ms/step - loss: 8.6840 - sparse_categorical_accuracy: 0.7763 - val_loss: 8.9303 - val_sparse_categorical_accuracy: 0.7754\n",
      "Epoch 199/300\n",
      "1/1 [==============================] - 0s 288ms/step - loss: 8.6582 - sparse_categorical_accuracy: 0.7763 - val_loss: 8.9088 - val_sparse_categorical_accuracy: 0.7755\n",
      "Epoch 200/300\n",
      "1/1 [==============================] - 0s 241ms/step - loss: 8.6326 - sparse_categorical_accuracy: 0.7764 - val_loss: 8.8875 - val_sparse_categorical_accuracy: 0.7762\n",
      "Epoch 201/300\n",
      "1/1 [==============================] - 0s 248ms/step - loss: 8.6072 - sparse_categorical_accuracy: 0.7768 - val_loss: 8.8663 - val_sparse_categorical_accuracy: 0.7761\n",
      "Epoch 202/300\n",
      "1/1 [==============================] - 0s 247ms/step - loss: 8.5818 - sparse_categorical_accuracy: 0.7771 - val_loss: 8.8452 - val_sparse_categorical_accuracy: 0.7766\n",
      "Epoch 203/300\n",
      "1/1 [==============================] - 0s 277ms/step - loss: 8.5567 - sparse_categorical_accuracy: 0.7772 - val_loss: 8.8242 - val_sparse_categorical_accuracy: 0.7763\n",
      "Epoch 204/300\n",
      "1/1 [==============================] - 0s 254ms/step - loss: 8.5317 - sparse_categorical_accuracy: 0.7774 - val_loss: 8.8034 - val_sparse_categorical_accuracy: 0.7764\n",
      "Epoch 205/300\n",
      "1/1 [==============================] - 0s 238ms/step - loss: 8.5068 - sparse_categorical_accuracy: 0.7778 - val_loss: 8.7828 - val_sparse_categorical_accuracy: 0.7765\n",
      "Epoch 206/300\n",
      "1/1 [==============================] - 0s 246ms/step - loss: 8.4820 - sparse_categorical_accuracy: 0.7780 - val_loss: 8.7622 - val_sparse_categorical_accuracy: 0.7764\n",
      "Epoch 207/300\n",
      "1/1 [==============================] - 0s 264ms/step - loss: 8.4574 - sparse_categorical_accuracy: 0.7784 - val_loss: 8.7417 - val_sparse_categorical_accuracy: 0.7767\n",
      "Epoch 208/300\n",
      "1/1 [==============================] - 0s 310ms/step - loss: 8.4330 - sparse_categorical_accuracy: 0.7785 - val_loss: 8.7214 - val_sparse_categorical_accuracy: 0.7770\n",
      "Epoch 209/300\n",
      "1/1 [==============================] - 0s 291ms/step - loss: 8.4086 - sparse_categorical_accuracy: 0.7786 - val_loss: 8.7011 - val_sparse_categorical_accuracy: 0.7776\n",
      "Epoch 210/300\n",
      "1/1 [==============================] - 0s 310ms/step - loss: 8.3844 - sparse_categorical_accuracy: 0.7788 - val_loss: 8.6810 - val_sparse_categorical_accuracy: 0.7780\n",
      "Epoch 211/300\n",
      "1/1 [==============================] - 0s 317ms/step - loss: 8.3604 - sparse_categorical_accuracy: 0.7790 - val_loss: 8.6610 - val_sparse_categorical_accuracy: 0.7780\n",
      "Epoch 212/300\n",
      "1/1 [==============================] - 0s 240ms/step - loss: 8.3365 - sparse_categorical_accuracy: 0.7793 - val_loss: 8.6411 - val_sparse_categorical_accuracy: 0.7781\n",
      "Epoch 213/300\n",
      "1/1 [==============================] - 0s 247ms/step - loss: 8.3127 - sparse_categorical_accuracy: 0.7793 - val_loss: 8.6213 - val_sparse_categorical_accuracy: 0.7784\n",
      "Epoch 214/300\n",
      "1/1 [==============================] - 0s 243ms/step - loss: 8.2890 - sparse_categorical_accuracy: 0.7794 - val_loss: 8.6017 - val_sparse_categorical_accuracy: 0.7786\n",
      "Epoch 215/300\n",
      "1/1 [==============================] - 0s 219ms/step - loss: 8.2654 - sparse_categorical_accuracy: 0.7795 - val_loss: 8.5821 - val_sparse_categorical_accuracy: 0.7786\n",
      "Epoch 216/300\n",
      "1/1 [==============================] - 0s 250ms/step - loss: 8.2420 - sparse_categorical_accuracy: 0.7797 - val_loss: 8.5627 - val_sparse_categorical_accuracy: 0.7789\n",
      "Epoch 217/300\n",
      "1/1 [==============================] - 0s 234ms/step - loss: 8.2187 - sparse_categorical_accuracy: 0.7798 - val_loss: 8.5434 - val_sparse_categorical_accuracy: 0.7787\n",
      "Epoch 218/300\n",
      "1/1 [==============================] - 0s 252ms/step - loss: 8.1955 - sparse_categorical_accuracy: 0.7800 - val_loss: 8.5242 - val_sparse_categorical_accuracy: 0.7790\n",
      "Epoch 219/300\n",
      "1/1 [==============================] - 0s 243ms/step - loss: 8.1724 - sparse_categorical_accuracy: 0.7799 - val_loss: 8.5050 - val_sparse_categorical_accuracy: 0.7791\n",
      "Epoch 220/300\n",
      "1/1 [==============================] - 0s 257ms/step - loss: 8.1494 - sparse_categorical_accuracy: 0.7802 - val_loss: 8.4860 - val_sparse_categorical_accuracy: 0.7794\n",
      "Epoch 221/300\n",
      "1/1 [==============================] - 0s 230ms/step - loss: 8.1266 - sparse_categorical_accuracy: 0.7803 - val_loss: 8.4671 - val_sparse_categorical_accuracy: 0.7793\n",
      "Epoch 222/300\n",
      "1/1 [==============================] - 0s 242ms/step - loss: 8.1039 - sparse_categorical_accuracy: 0.7806 - val_loss: 8.4482 - val_sparse_categorical_accuracy: 0.7795\n",
      "Epoch 223/300\n",
      "1/1 [==============================] - 0s 242ms/step - loss: 8.0812 - sparse_categorical_accuracy: 0.7807 - val_loss: 8.4295 - val_sparse_categorical_accuracy: 0.7796\n",
      "Epoch 224/300\n",
      "1/1 [==============================] - 0s 249ms/step - loss: 8.0588 - sparse_categorical_accuracy: 0.7807 - val_loss: 8.4108 - val_sparse_categorical_accuracy: 0.7798\n",
      "Epoch 225/300\n",
      "1/1 [==============================] - 0s 237ms/step - loss: 8.0364 - sparse_categorical_accuracy: 0.7809 - val_loss: 8.3922 - val_sparse_categorical_accuracy: 0.7798\n",
      "Epoch 226/300\n",
      "1/1 [==============================] - 0s 241ms/step - loss: 8.0141 - sparse_categorical_accuracy: 0.7811 - val_loss: 8.3737 - val_sparse_categorical_accuracy: 0.7798\n",
      "Epoch 227/300\n",
      "1/1 [==============================] - 0s 235ms/step - loss: 7.9920 - sparse_categorical_accuracy: 0.7812 - val_loss: 8.3553 - val_sparse_categorical_accuracy: 0.7795\n",
      "Epoch 228/300\n",
      "1/1 [==============================] - 0s 248ms/step - loss: 7.9700 - sparse_categorical_accuracy: 0.7814 - val_loss: 8.3370 - val_sparse_categorical_accuracy: 0.7794\n",
      "Epoch 229/300\n",
      "1/1 [==============================] - 0s 258ms/step - loss: 7.9480 - sparse_categorical_accuracy: 0.7816 - val_loss: 8.3189 - val_sparse_categorical_accuracy: 0.7792\n",
      "Epoch 230/300\n",
      "1/1 [==============================] - 0s 241ms/step - loss: 7.9262 - sparse_categorical_accuracy: 0.7817 - val_loss: 8.3008 - val_sparse_categorical_accuracy: 0.7790\n",
      "Epoch 231/300\n",
      "1/1 [==============================] - 0s 243ms/step - loss: 7.9046 - sparse_categorical_accuracy: 0.7818 - val_loss: 8.2829 - val_sparse_categorical_accuracy: 0.7789\n",
      "Epoch 232/300\n",
      "1/1 [==============================] - 0s 229ms/step - loss: 7.8830 - sparse_categorical_accuracy: 0.7819 - val_loss: 8.2651 - val_sparse_categorical_accuracy: 0.7792\n",
      "Epoch 233/300\n",
      "1/1 [==============================] - 0s 241ms/step - loss: 7.8615 - sparse_categorical_accuracy: 0.7821 - val_loss: 8.2474 - val_sparse_categorical_accuracy: 0.7793\n",
      "Epoch 234/300\n",
      "1/1 [==============================] - 0s 237ms/step - loss: 7.8402 - sparse_categorical_accuracy: 0.7822 - val_loss: 8.2298 - val_sparse_categorical_accuracy: 0.7793\n",
      "Epoch 235/300\n",
      "1/1 [==============================] - 0s 242ms/step - loss: 7.8189 - sparse_categorical_accuracy: 0.7824 - val_loss: 8.2123 - val_sparse_categorical_accuracy: 0.7795\n",
      "Epoch 236/300\n",
      "1/1 [==============================] - 0s 245ms/step - loss: 7.7978 - sparse_categorical_accuracy: 0.7825 - val_loss: 8.1949 - val_sparse_categorical_accuracy: 0.7791\n",
      "Epoch 237/300\n",
      "1/1 [==============================] - 0s 249ms/step - loss: 7.7768 - sparse_categorical_accuracy: 0.7828 - val_loss: 8.1776 - val_sparse_categorical_accuracy: 0.7791\n",
      "Epoch 238/300\n",
      "1/1 [==============================] - 0s 265ms/step - loss: 7.7559 - sparse_categorical_accuracy: 0.7830 - val_loss: 8.1603 - val_sparse_categorical_accuracy: 0.7792\n",
      "Epoch 239/300\n",
      "1/1 [==============================] - 0s 285ms/step - loss: 7.7350 - sparse_categorical_accuracy: 0.7832 - val_loss: 8.1432 - val_sparse_categorical_accuracy: 0.7791\n",
      "Epoch 240/300\n",
      "1/1 [==============================] - 0s 263ms/step - loss: 7.7143 - sparse_categorical_accuracy: 0.7833 - val_loss: 8.1261 - val_sparse_categorical_accuracy: 0.7793\n",
      "Epoch 241/300\n",
      "1/1 [==============================] - 0s 247ms/step - loss: 7.6937 - sparse_categorical_accuracy: 0.7836 - val_loss: 8.1090 - val_sparse_categorical_accuracy: 0.7795\n",
      "Epoch 242/300\n",
      "1/1 [==============================] - 0s 254ms/step - loss: 7.6732 - sparse_categorical_accuracy: 0.7837 - val_loss: 8.0921 - val_sparse_categorical_accuracy: 0.7797\n",
      "Epoch 243/300\n",
      "1/1 [==============================] - 0s 257ms/step - loss: 7.6528 - sparse_categorical_accuracy: 0.7838 - val_loss: 8.0752 - val_sparse_categorical_accuracy: 0.7800\n",
      "Epoch 244/300\n",
      "1/1 [==============================] - 0s 248ms/step - loss: 7.6324 - sparse_categorical_accuracy: 0.7839 - val_loss: 8.0583 - val_sparse_categorical_accuracy: 0.7805\n",
      "Epoch 245/300\n",
      "1/1 [==============================] - 0s 233ms/step - loss: 7.6122 - sparse_categorical_accuracy: 0.7841 - val_loss: 8.0416 - val_sparse_categorical_accuracy: 0.7807\n",
      "Epoch 246/300\n",
      "1/1 [==============================] - 0s 232ms/step - loss: 7.5921 - sparse_categorical_accuracy: 0.7840 - val_loss: 8.0248 - val_sparse_categorical_accuracy: 0.7811\n",
      "Epoch 247/300\n",
      "1/1 [==============================] - 0s 252ms/step - loss: 7.5720 - sparse_categorical_accuracy: 0.7841 - val_loss: 8.0082 - val_sparse_categorical_accuracy: 0.7813\n",
      "Epoch 248/300\n",
      "1/1 [==============================] - 0s 244ms/step - loss: 7.5521 - sparse_categorical_accuracy: 0.7843 - val_loss: 7.9916 - val_sparse_categorical_accuracy: 0.7817\n",
      "Epoch 249/300\n",
      "1/1 [==============================] - 0s 264ms/step - loss: 7.5322 - sparse_categorical_accuracy: 0.7844 - val_loss: 7.9751 - val_sparse_categorical_accuracy: 0.7818\n",
      "Epoch 250/300\n",
      "1/1 [==============================] - 0s 246ms/step - loss: 7.5124 - sparse_categorical_accuracy: 0.7847 - val_loss: 7.9586 - val_sparse_categorical_accuracy: 0.7820\n",
      "Epoch 251/300\n",
      "1/1 [==============================] - 0s 270ms/step - loss: 7.4928 - sparse_categorical_accuracy: 0.7848 - val_loss: 7.9423 - val_sparse_categorical_accuracy: 0.7822\n",
      "Epoch 252/300\n",
      "1/1 [==============================] - 0s 249ms/step - loss: 7.4732 - sparse_categorical_accuracy: 0.7850 - val_loss: 7.9260 - val_sparse_categorical_accuracy: 0.7820\n",
      "Epoch 253/300\n",
      "1/1 [==============================] - 0s 244ms/step - loss: 7.4536 - sparse_categorical_accuracy: 0.7850 - val_loss: 7.9098 - val_sparse_categorical_accuracy: 0.7821\n",
      "Epoch 254/300\n",
      "1/1 [==============================] - 0s 234ms/step - loss: 7.4342 - sparse_categorical_accuracy: 0.7853 - val_loss: 7.8937 - val_sparse_categorical_accuracy: 0.7820\n",
      "Epoch 255/300\n",
      "1/1 [==============================] - 0s 238ms/step - loss: 7.4149 - sparse_categorical_accuracy: 0.7854 - val_loss: 7.8777 - val_sparse_categorical_accuracy: 0.7819\n",
      "Epoch 256/300\n",
      "1/1 [==============================] - 0s 250ms/step - loss: 7.3957 - sparse_categorical_accuracy: 0.7856 - val_loss: 7.8617 - val_sparse_categorical_accuracy: 0.7819\n",
      "Epoch 257/300\n",
      "1/1 [==============================] - 0s 255ms/step - loss: 7.3765 - sparse_categorical_accuracy: 0.7858 - val_loss: 7.8458 - val_sparse_categorical_accuracy: 0.7821\n",
      "Epoch 258/300\n",
      "1/1 [==============================] - 0s 240ms/step - loss: 7.3575 - sparse_categorical_accuracy: 0.7860 - val_loss: 7.8300 - val_sparse_categorical_accuracy: 0.7826\n",
      "Epoch 259/300\n",
      "1/1 [==============================] - 0s 241ms/step - loss: 7.3385 - sparse_categorical_accuracy: 0.7861 - val_loss: 7.8143 - val_sparse_categorical_accuracy: 0.7831\n",
      "Epoch 260/300\n",
      "1/1 [==============================] - 0s 245ms/step - loss: 7.3196 - sparse_categorical_accuracy: 0.7862 - val_loss: 7.7986 - val_sparse_categorical_accuracy: 0.7829\n",
      "Epoch 261/300\n",
      "1/1 [==============================] - 0s 285ms/step - loss: 7.3008 - sparse_categorical_accuracy: 0.7864 - val_loss: 7.7830 - val_sparse_categorical_accuracy: 0.7831\n",
      "Epoch 262/300\n",
      "1/1 [==============================] - 0s 251ms/step - loss: 7.2821 - sparse_categorical_accuracy: 0.7866 - val_loss: 7.7675 - val_sparse_categorical_accuracy: 0.7833\n",
      "Epoch 263/300\n",
      "1/1 [==============================] - 0s 317ms/step - loss: 7.2635 - sparse_categorical_accuracy: 0.7869 - val_loss: 7.7521 - val_sparse_categorical_accuracy: 0.7835\n",
      "Epoch 264/300\n",
      "1/1 [==============================] - 0s 245ms/step - loss: 7.2450 - sparse_categorical_accuracy: 0.7872 - val_loss: 7.7367 - val_sparse_categorical_accuracy: 0.7839\n",
      "Epoch 265/300\n",
      "1/1 [==============================] - 0s 253ms/step - loss: 7.2265 - sparse_categorical_accuracy: 0.7874 - val_loss: 7.7213 - val_sparse_categorical_accuracy: 0.7841\n",
      "Epoch 266/300\n",
      "1/1 [==============================] - 0s 294ms/step - loss: 7.2081 - sparse_categorical_accuracy: 0.7875 - val_loss: 7.7061 - val_sparse_categorical_accuracy: 0.7843\n",
      "Epoch 267/300\n",
      "1/1 [==============================] - 0s 241ms/step - loss: 7.1899 - sparse_categorical_accuracy: 0.7878 - val_loss: 7.6909 - val_sparse_categorical_accuracy: 0.7844\n",
      "Epoch 268/300\n",
      "1/1 [==============================] - 0s 249ms/step - loss: 7.1717 - sparse_categorical_accuracy: 0.7879 - val_loss: 7.6757 - val_sparse_categorical_accuracy: 0.7846\n",
      "Epoch 269/300\n",
      "1/1 [==============================] - 0s 250ms/step - loss: 7.1536 - sparse_categorical_accuracy: 0.7879 - val_loss: 7.6607 - val_sparse_categorical_accuracy: 0.7851\n",
      "Epoch 270/300\n",
      "1/1 [==============================] - 0s 249ms/step - loss: 7.1355 - sparse_categorical_accuracy: 0.7880 - val_loss: 7.6457 - val_sparse_categorical_accuracy: 0.7851\n",
      "Epoch 271/300\n",
      "1/1 [==============================] - 0s 254ms/step - loss: 7.1176 - sparse_categorical_accuracy: 0.7881 - val_loss: 7.6308 - val_sparse_categorical_accuracy: 0.7852\n",
      "Epoch 272/300\n",
      "1/1 [==============================] - 0s 226ms/step - loss: 7.0997 - sparse_categorical_accuracy: 0.7883 - val_loss: 7.6160 - val_sparse_categorical_accuracy: 0.7856\n",
      "Epoch 273/300\n",
      "1/1 [==============================] - 0s 239ms/step - loss: 7.0819 - sparse_categorical_accuracy: 0.7885 - val_loss: 7.6013 - val_sparse_categorical_accuracy: 0.7855\n",
      "Epoch 274/300\n",
      "1/1 [==============================] - 0s 229ms/step - loss: 7.0642 - sparse_categorical_accuracy: 0.7886 - val_loss: 7.5867 - val_sparse_categorical_accuracy: 0.7858\n",
      "Epoch 275/300\n",
      "1/1 [==============================] - 0s 236ms/step - loss: 7.0465 - sparse_categorical_accuracy: 0.7886 - val_loss: 7.5721 - val_sparse_categorical_accuracy: 0.7859\n",
      "Epoch 276/300\n",
      "1/1 [==============================] - 0s 248ms/step - loss: 7.0290 - sparse_categorical_accuracy: 0.7886 - val_loss: 7.5576 - val_sparse_categorical_accuracy: 0.7862\n",
      "Epoch 277/300\n",
      "1/1 [==============================] - 0s 237ms/step - loss: 7.0115 - sparse_categorical_accuracy: 0.7888 - val_loss: 7.5432 - val_sparse_categorical_accuracy: 0.7863\n",
      "Epoch 278/300\n",
      "1/1 [==============================] - 0s 250ms/step - loss: 6.9941 - sparse_categorical_accuracy: 0.7891 - val_loss: 7.5288 - val_sparse_categorical_accuracy: 0.7862\n",
      "Epoch 279/300\n",
      "1/1 [==============================] - 0s 225ms/step - loss: 6.9767 - sparse_categorical_accuracy: 0.7893 - val_loss: 7.5145 - val_sparse_categorical_accuracy: 0.7863\n",
      "Epoch 280/300\n",
      "1/1 [==============================] - 0s 251ms/step - loss: 6.9595 - sparse_categorical_accuracy: 0.7893 - val_loss: 7.5002 - val_sparse_categorical_accuracy: 0.7863\n",
      "Epoch 281/300\n",
      "1/1 [==============================] - 0s 247ms/step - loss: 6.9423 - sparse_categorical_accuracy: 0.7895 - val_loss: 7.4860 - val_sparse_categorical_accuracy: 0.7864\n",
      "Epoch 282/300\n",
      "1/1 [==============================] - 0s 251ms/step - loss: 6.9252 - sparse_categorical_accuracy: 0.7895 - val_loss: 7.4719 - val_sparse_categorical_accuracy: 0.7866\n",
      "Epoch 283/300\n",
      "1/1 [==============================] - 0s 264ms/step - loss: 6.9081 - sparse_categorical_accuracy: 0.7895 - val_loss: 7.4578 - val_sparse_categorical_accuracy: 0.7867\n",
      "Epoch 284/300\n",
      "1/1 [==============================] - 0s 243ms/step - loss: 6.8912 - sparse_categorical_accuracy: 0.7895 - val_loss: 7.4438 - val_sparse_categorical_accuracy: 0.7862\n",
      "Epoch 285/300\n",
      "1/1 [==============================] - 0s 284ms/step - loss: 6.8743 - sparse_categorical_accuracy: 0.7897 - val_loss: 7.4299 - val_sparse_categorical_accuracy: 0.7863\n",
      "Epoch 286/300\n",
      "1/1 [==============================] - 0s 248ms/step - loss: 6.8575 - sparse_categorical_accuracy: 0.7898 - val_loss: 7.4160 - val_sparse_categorical_accuracy: 0.7867\n",
      "Epoch 287/300\n",
      "1/1 [==============================] - 0s 237ms/step - loss: 6.8408 - sparse_categorical_accuracy: 0.7900 - val_loss: 7.4022 - val_sparse_categorical_accuracy: 0.7867\n",
      "Epoch 288/300\n",
      "1/1 [==============================] - 0s 247ms/step - loss: 6.8241 - sparse_categorical_accuracy: 0.7901 - val_loss: 7.3885 - val_sparse_categorical_accuracy: 0.7869\n",
      "Epoch 289/300\n",
      "1/1 [==============================] - 0s 232ms/step - loss: 6.8075 - sparse_categorical_accuracy: 0.7901 - val_loss: 7.3749 - val_sparse_categorical_accuracy: 0.7869\n",
      "Epoch 290/300\n",
      "1/1 [==============================] - 0s 244ms/step - loss: 6.7910 - sparse_categorical_accuracy: 0.7902 - val_loss: 7.3614 - val_sparse_categorical_accuracy: 0.7871\n",
      "Epoch 291/300\n",
      "1/1 [==============================] - 0s 247ms/step - loss: 6.7746 - sparse_categorical_accuracy: 0.7902 - val_loss: 7.3479 - val_sparse_categorical_accuracy: 0.7873\n",
      "Epoch 292/300\n",
      "1/1 [==============================] - 0s 249ms/step - loss: 6.7582 - sparse_categorical_accuracy: 0.7904 - val_loss: 7.3346 - val_sparse_categorical_accuracy: 0.7875\n",
      "Epoch 293/300\n",
      "1/1 [==============================] - 0s 239ms/step - loss: 6.7419 - sparse_categorical_accuracy: 0.7904 - val_loss: 7.3213 - val_sparse_categorical_accuracy: 0.7874\n",
      "Epoch 294/300\n",
      "1/1 [==============================] - 0s 249ms/step - loss: 6.7257 - sparse_categorical_accuracy: 0.7906 - val_loss: 7.3081 - val_sparse_categorical_accuracy: 0.7876\n",
      "Epoch 295/300\n",
      "1/1 [==============================] - 0s 248ms/step - loss: 6.7095 - sparse_categorical_accuracy: 0.7907 - val_loss: 7.2949 - val_sparse_categorical_accuracy: 0.7881\n",
      "Epoch 296/300\n",
      "1/1 [==============================] - 0s 259ms/step - loss: 6.6934 - sparse_categorical_accuracy: 0.7908 - val_loss: 7.2819 - val_sparse_categorical_accuracy: 0.7881\n",
      "Epoch 297/300\n",
      "1/1 [==============================] - 0s 185ms/step - loss: 6.6774 - sparse_categorical_accuracy: 0.7908 - val_loss: 7.2688 - val_sparse_categorical_accuracy: 0.7883\n",
      "Epoch 298/300\n",
      "1/1 [==============================] - 0s 168ms/step - loss: 6.6615 - sparse_categorical_accuracy: 0.7909 - val_loss: 7.2559 - val_sparse_categorical_accuracy: 0.7885\n",
      "Epoch 299/300\n",
      "1/1 [==============================] - 0s 169ms/step - loss: 6.6456 - sparse_categorical_accuracy: 0.7911 - val_loss: 7.2429 - val_sparse_categorical_accuracy: 0.7887\n",
      "Epoch 300/300\n",
      "1/1 [==============================] - 0s 173ms/step - loss: 6.6297 - sparse_categorical_accuracy: 0.7912 - val_loss: 7.2301 - val_sparse_categorical_accuracy: 0.7885\n"
     ]
    }
   ],
   "source": [
    "k_l2=0\n",
    "keras_model = tf.keras.Sequential([\n",
    "    #tf.keras.layers.Dense(20, activation='tanh',kernel_regularizer=keras.regularizers.l2(k_l2)),\n",
    "    #tf.keras.layers.Dense(20, activation='tanh',kernel_regularizer=keras.regularizers.l2(k_l2)),\n",
    "    tf.keras.layers.Dense(10, activation='softmax', kernel_regularizer=keras.regularizers.l2(k_l2))\n",
    "])\n",
    "\n",
    "keras_model.build(input_shape=[None, 784])\n",
    "keras_model.summary()\n",
    "\n",
    "keras_model.compile(\n",
    "    optimizer=tf.keras.optimizers.Adam(), # Optimizer\n",
    "    loss=tf.keras.losses.SparseCategoricalCrossentropy(), # Loss function to minimize\n",
    "    metrics=[keras.metrics.SparseCategoricalAccuracy()] # List of metrics to monitor\n",
    ")\n",
    "\n",
    "history = keras_model.fit(\n",
    "    train_input,\n",
    "    train_label,\n",
    "    batch_size=len(train_input),\n",
    "    epochs=300,\n",
    "    validation_data=(test_input, test_label)\n",
    ")\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 90,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Key: loss\n",
      "Key: sparse_categorical_accuracy\n",
      "Key: val_loss\n",
      "Key: val_sparse_categorical_accuracy\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# vizualize\n",
    "\n",
    "for item in history.history:\n",
    "    print(\"Key:\",item)\n",
    "\n",
    "plt.plot(history.history['loss'],label=\"train\")\n",
    "plt.plot(history.history['val_loss'],label=\"validation\")\n",
    "plt.title('Model Loss')\n",
    "plt.yscale('log')\n",
    "plt.ylabel('Cross Entropy')\n",
    "plt.xlabel('Iteration')\n",
    "plt.grid()\n",
    "plt.legend()\n",
    "plt.show()\n",
    "\n",
    "#print(\"history\",history.history)\n",
    "plt.plot(history.history['sparse_categorical_accuracy'],label=\"train\")\n",
    "plt.plot(history.history['val_sparse_categorical_accuracy'],label=\"validation\")\n",
    "plt.title('Model Accuracy')\n",
    "#plt.yscale('log')\n",
    "plt.ylabel('Acc')\n",
    "plt.xlabel('Iteration')\n",
    "plt.grid()\n",
    "plt.legend()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Lab31"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 2.\n",
    "\n",
    "Adaptuokite Lab24 MNIST kodą pridedant konvoliucinius sluoksnius (MaxPooling2D, Conv2D, Flatten)."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model: \"sequential_17\"\n",
      "_________________________________________________________________\n",
      " Layer (type)                Output Shape              Param #   \n",
      "=================================================================\n",
      " conv2d_24 (Conv2D)          (None, 27, 27, 8)         40        \n",
      "                                                                 \n",
      " max_pooling2d_21 (MaxPooli  (None, 13, 13, 8)         0         \n",
      " ng2D)                                                           \n",
      "                                                                 \n",
      " conv2d_25 (Conv2D)          (None, 12, 12, 16)        528       \n",
      "                                                                 \n",
      " max_pooling2d_22 (MaxPooli  (None, 6, 6, 16)          0         \n",
      " ng2D)                                                           \n",
      "                                                                 \n",
      " flatten_16 (Flatten)        (None, 576)               0         \n",
      "                                                                 \n",
      " dense_31 (Dense)            (None, 10)                5770      \n",
      "                                                                 \n",
      "=================================================================\n",
      "Total params: 6338 (24.76 KB)\n",
      "Trainable params: 6338 (24.76 KB)\n",
      "Non-trainable params: 0 (0.00 Byte)\n",
      "_________________________________________________________________\n",
      "Epoch 1/200\n",
      "1/1 [==============================] - 11s 11s/step - loss: 2.3231 - sparse_categorical_accuracy: 0.0935 - val_loss: 2.3027 - val_sparse_categorical_accuracy: 0.1181\n",
      "Epoch 2/200\n",
      "1/1 [==============================] - 2s 2s/step - loss: 2.3031 - sparse_categorical_accuracy: 0.1187 - val_loss: 2.2867 - val_sparse_categorical_accuracy: 0.1464\n",
      "Epoch 3/200\n",
      "1/1 [==============================] - 2s 2s/step - loss: 2.2872 - sparse_categorical_accuracy: 0.1497 - val_loss: 2.2725 - val_sparse_categorical_accuracy: 0.1778\n",
      "Epoch 4/200\n",
      "1/1 [==============================] - 2s 2s/step - loss: 2.2730 - sparse_categorical_accuracy: 0.1830 - val_loss: 2.2586 - val_sparse_categorical_accuracy: 0.2101\n",
      "Epoch 5/200\n",
      "1/1 [==============================] - 2s 2s/step - loss: 2.2593 - sparse_categorical_accuracy: 0.2157 - val_loss: 2.2436 - val_sparse_categorical_accuracy: 0.2497\n",
      "Epoch 6/200\n",
      "1/1 [==============================] - 2s 2s/step - loss: 2.2444 - sparse_categorical_accuracy: 0.2482 - val_loss: 2.2265 - val_sparse_categorical_accuracy: 0.2922\n",
      "Epoch 7/200\n",
      "1/1 [==============================] - 2s 2s/step - loss: 2.2275 - sparse_categorical_accuracy: 0.2905 - val_loss: 2.2064 - val_sparse_categorical_accuracy: 0.3411\n",
      "Epoch 8/200\n",
      "1/1 [==============================] - 2s 2s/step - loss: 2.2076 - sparse_categorical_accuracy: 0.3380 - val_loss: 2.1813 - val_sparse_categorical_accuracy: 0.3924\n",
      "Epoch 9/200\n",
      "1/1 [==============================] - 2s 2s/step - loss: 2.1829 - sparse_categorical_accuracy: 0.3882 - val_loss: 2.1496 - val_sparse_categorical_accuracy: 0.4492\n",
      "Epoch 10/200\n",
      "1/1 [==============================] - 2s 2s/step - loss: 2.1515 - sparse_categorical_accuracy: 0.4465 - val_loss: 2.1091 - val_sparse_categorical_accuracy: 0.5022\n",
      "Epoch 11/200\n",
      "1/1 [==============================] - 2s 2s/step - loss: 2.1115 - sparse_categorical_accuracy: 0.4981 - val_loss: 2.0564 - val_sparse_categorical_accuracy: 0.5529\n",
      "Epoch 12/200\n",
      "1/1 [==============================] - 2s 2s/step - loss: 2.0595 - sparse_categorical_accuracy: 0.5463 - val_loss: 1.9875 - val_sparse_categorical_accuracy: 0.5811\n",
      "Epoch 13/200\n",
      "1/1 [==============================] - 2s 2s/step - loss: 1.9916 - sparse_categorical_accuracy: 0.5745 - val_loss: 1.8943 - val_sparse_categorical_accuracy: 0.6322\n",
      "Epoch 14/200\n",
      "1/1 [==============================] - 2s 2s/step - loss: 1.8999 - sparse_categorical_accuracy: 0.6226 - val_loss: 1.7726 - val_sparse_categorical_accuracy: 0.6620\n",
      "Epoch 15/200\n",
      "1/1 [==============================] - 2s 2s/step - loss: 1.7801 - sparse_categorical_accuracy: 0.6516 - val_loss: 1.6186 - val_sparse_categorical_accuracy: 0.6927\n",
      "Epoch 16/200\n",
      "1/1 [==============================] - 2s 2s/step - loss: 1.6289 - sparse_categorical_accuracy: 0.6790 - val_loss: 1.4365 - val_sparse_categorical_accuracy: 0.7234\n",
      "Epoch 17/200\n",
      "1/1 [==============================] - 2s 2s/step - loss: 1.4507 - sparse_categorical_accuracy: 0.7081 - val_loss: 1.2450 - val_sparse_categorical_accuracy: 0.7447\n",
      "Epoch 18/200\n",
      "1/1 [==============================] - 2s 2s/step - loss: 1.2634 - sparse_categorical_accuracy: 0.7287 - val_loss: 1.0771 - val_sparse_categorical_accuracy: 0.7580\n",
      "Epoch 19/200\n",
      "1/1 [==============================] - 2s 2s/step - loss: 1.1001 - sparse_categorical_accuracy: 0.7435 - val_loss: 1.0968 - val_sparse_categorical_accuracy: 0.6258\n",
      "Epoch 20/200\n",
      "1/1 [==============================] - 2s 2s/step - loss: 1.1150 - sparse_categorical_accuracy: 0.6191 - val_loss: 2.8059 - val_sparse_categorical_accuracy: 0.4252\n",
      "Epoch 21/200\n",
      "1/1 [==============================] - 2s 2s/step - loss: 2.8337 - sparse_categorical_accuracy: 0.4113 - val_loss: 2.7845 - val_sparse_categorical_accuracy: 0.4922\n",
      "Epoch 22/200\n",
      "1/1 [==============================] - 2s 2s/step - loss: 2.7802 - sparse_categorical_accuracy: 0.4894 - val_loss: 1.9623 - val_sparse_categorical_accuracy: 0.4897\n",
      "Epoch 23/200\n",
      "1/1 [==============================] - 2s 2s/step - loss: 1.9748 - sparse_categorical_accuracy: 0.4824 - val_loss: 1.8047 - val_sparse_categorical_accuracy: 0.6128\n",
      "Epoch 24/200\n",
      "1/1 [==============================] - 2s 2s/step - loss: 1.8204 - sparse_categorical_accuracy: 0.6014 - val_loss: 1.6263 - val_sparse_categorical_accuracy: 0.6479\n",
      "Epoch 25/200\n",
      "1/1 [==============================] - 2s 2s/step - loss: 1.6444 - sparse_categorical_accuracy: 0.6369 - val_loss: 1.3955 - val_sparse_categorical_accuracy: 0.6670\n",
      "Epoch 26/200\n",
      "1/1 [==============================] - 2s 2s/step - loss: 1.4163 - sparse_categorical_accuracy: 0.6573 - val_loss: 1.1135 - val_sparse_categorical_accuracy: 0.7711\n",
      "Epoch 27/200\n",
      "1/1 [==============================] - 2s 2s/step - loss: 1.1382 - sparse_categorical_accuracy: 0.7553 - val_loss: 0.8681 - val_sparse_categorical_accuracy: 0.8239\n",
      "Epoch 28/200\n",
      "1/1 [==============================] - 2s 2s/step - loss: 0.8970 - sparse_categorical_accuracy: 0.8044 - val_loss: 0.7336 - val_sparse_categorical_accuracy: 0.8258\n",
      "Epoch 29/200\n",
      "1/1 [==============================] - 2s 2s/step - loss: 0.7629 - sparse_categorical_accuracy: 0.8118 - val_loss: 0.6685 - val_sparse_categorical_accuracy: 0.8254\n",
      "Epoch 30/200\n",
      "1/1 [==============================] - 2s 2s/step - loss: 0.7002 - sparse_categorical_accuracy: 0.8078 - val_loss: 0.7937 - val_sparse_categorical_accuracy: 0.7718\n",
      "Epoch 31/200\n",
      "1/1 [==============================] - 2s 2s/step - loss: 0.8176 - sparse_categorical_accuracy: 0.7602 - val_loss: 1.4872 - val_sparse_categorical_accuracy: 0.4954\n",
      "Epoch 32/200\n",
      "1/1 [==============================] - 2s 2s/step - loss: 1.5201 - sparse_categorical_accuracy: 0.4868 - val_loss: 2.7679 - val_sparse_categorical_accuracy: 0.6833\n",
      "Epoch 33/200\n",
      "1/1 [==============================] - 2s 2s/step - loss: 2.7496 - sparse_categorical_accuracy: 0.6783 - val_loss: 1.7345 - val_sparse_categorical_accuracy: 0.6377\n",
      "Epoch 34/200\n",
      "1/1 [==============================] - 2s 2s/step - loss: 1.7398 - sparse_categorical_accuracy: 0.6330 - val_loss: 1.5192 - val_sparse_categorical_accuracy: 0.7493\n",
      "Epoch 35/200\n",
      "1/1 [==============================] - 2s 2s/step - loss: 1.5273 - sparse_categorical_accuracy: 0.7374 - val_loss: 1.3959 - val_sparse_categorical_accuracy: 0.7688\n",
      "Epoch 36/200\n",
      "1/1 [==============================] - 2s 2s/step - loss: 1.4055 - sparse_categorical_accuracy: 0.7601 - val_loss: 1.2664 - val_sparse_categorical_accuracy: 0.7812\n",
      "Epoch 37/200\n",
      "1/1 [==============================] - 2s 2s/step - loss: 1.2776 - sparse_categorical_accuracy: 0.7734 - val_loss: 1.1224 - val_sparse_categorical_accuracy: 0.7852\n",
      "Epoch 38/200\n",
      "1/1 [==============================] - 2s 2s/step - loss: 1.1353 - sparse_categorical_accuracy: 0.7782 - val_loss: 0.9562 - val_sparse_categorical_accuracy: 0.7860\n",
      "Epoch 39/200\n",
      "1/1 [==============================] - 2s 2s/step - loss: 0.9709 - sparse_categorical_accuracy: 0.7802 - val_loss: 0.7709 - val_sparse_categorical_accuracy: 0.8039\n",
      "Epoch 40/200\n",
      "1/1 [==============================] - 2s 2s/step - loss: 0.7881 - sparse_categorical_accuracy: 0.7975 - val_loss: 0.6252 - val_sparse_categorical_accuracy: 0.8466\n",
      "Epoch 41/200\n",
      "1/1 [==============================] - 2s 2s/step - loss: 0.6471 - sparse_categorical_accuracy: 0.8396 - val_loss: 0.5564 - val_sparse_categorical_accuracy: 0.8529\n",
      "Epoch 42/200\n",
      "1/1 [==============================] - 2s 2s/step - loss: 0.5797 - sparse_categorical_accuracy: 0.8446 - val_loss: 0.5172 - val_sparse_categorical_accuracy: 0.8564\n",
      "Epoch 43/200\n",
      "1/1 [==============================] - 2s 2s/step - loss: 0.5409 - sparse_categorical_accuracy: 0.8483 - val_loss: 0.4909 - val_sparse_categorical_accuracy: 0.8621\n",
      "Epoch 44/200\n",
      "1/1 [==============================] - 2s 2s/step - loss: 0.5147 - sparse_categorical_accuracy: 0.8522 - val_loss: 0.4721 - val_sparse_categorical_accuracy: 0.8651\n",
      "Epoch 45/200\n",
      "1/1 [==============================] - 2s 2s/step - loss: 0.4960 - sparse_categorical_accuracy: 0.8548 - val_loss: 0.4612 - val_sparse_categorical_accuracy: 0.8674\n",
      "Epoch 46/200\n",
      "1/1 [==============================] - 2s 2s/step - loss: 0.4848 - sparse_categorical_accuracy: 0.8554 - val_loss: 0.4766 - val_sparse_categorical_accuracy: 0.8586\n",
      "Epoch 47/200\n",
      "1/1 [==============================] - 2s 2s/step - loss: 0.5004 - sparse_categorical_accuracy: 0.8462 - val_loss: 0.6405 - val_sparse_categorical_accuracy: 0.7917\n",
      "Epoch 48/200\n",
      "1/1 [==============================] - 2s 2s/step - loss: 0.6626 - sparse_categorical_accuracy: 0.7847 - val_loss: 1.2383 - val_sparse_categorical_accuracy: 0.6740\n",
      "Epoch 49/200\n",
      "1/1 [==============================] - 2s 2s/step - loss: 1.2615 - sparse_categorical_accuracy: 0.6662 - val_loss: 1.5462 - val_sparse_categorical_accuracy: 0.7564\n",
      "Epoch 50/200\n",
      "1/1 [==============================] - 2s 2s/step - loss: 1.5575 - sparse_categorical_accuracy: 0.7496 - val_loss: 0.9303 - val_sparse_categorical_accuracy: 0.7687\n",
      "Epoch 51/200\n",
      "1/1 [==============================] - 2s 2s/step - loss: 0.9450 - sparse_categorical_accuracy: 0.7586 - val_loss: 0.6835 - val_sparse_categorical_accuracy: 0.8314\n",
      "Epoch 52/200\n",
      "1/1 [==============================] - 2s 2s/step - loss: 0.7040 - sparse_categorical_accuracy: 0.8248 - val_loss: 0.5411 - val_sparse_categorical_accuracy: 0.8726\n",
      "Epoch 53/200\n",
      "1/1 [==============================] - 2s 2s/step - loss: 0.5638 - sparse_categorical_accuracy: 0.8659 - val_loss: 0.4692 - val_sparse_categorical_accuracy: 0.8820\n",
      "Epoch 54/200\n",
      "1/1 [==============================] - 2s 2s/step - loss: 0.4922 - sparse_categorical_accuracy: 0.8741 - val_loss: 0.4347 - val_sparse_categorical_accuracy: 0.8861\n",
      "Epoch 55/200\n",
      "1/1 [==============================] - 2s 2s/step - loss: 0.4576 - sparse_categorical_accuracy: 0.8764 - val_loss: 0.4143 - val_sparse_categorical_accuracy: 0.8863\n",
      "Epoch 56/200\n",
      "1/1 [==============================] - 2s 2s/step - loss: 0.4368 - sparse_categorical_accuracy: 0.8777 - val_loss: 0.4003 - val_sparse_categorical_accuracy: 0.8881\n",
      "Epoch 57/200\n",
      "1/1 [==============================] - 2s 2s/step - loss: 0.4225 - sparse_categorical_accuracy: 0.8794 - val_loss: 0.3901 - val_sparse_categorical_accuracy: 0.8890\n",
      "Epoch 58/200\n",
      "1/1 [==============================] - 2s 2s/step - loss: 0.4118 - sparse_categorical_accuracy: 0.8803 - val_loss: 0.3817 - val_sparse_categorical_accuracy: 0.8908\n",
      "Epoch 59/200\n",
      "1/1 [==============================] - 2s 2s/step - loss: 0.4032 - sparse_categorical_accuracy: 0.8826 - val_loss: 0.3750 - val_sparse_categorical_accuracy: 0.8920\n",
      "Epoch 60/200\n",
      "1/1 [==============================] - 2s 2s/step - loss: 0.3961 - sparse_categorical_accuracy: 0.8833 - val_loss: 0.3688 - val_sparse_categorical_accuracy: 0.8945\n",
      "Epoch 61/200\n",
      "1/1 [==============================] - 2s 2s/step - loss: 0.3898 - sparse_categorical_accuracy: 0.8855 - val_loss: 0.3639 - val_sparse_categorical_accuracy: 0.8938\n",
      "Epoch 62/200\n",
      "1/1 [==============================] - 2s 2s/step - loss: 0.3843 - sparse_categorical_accuracy: 0.8856 - val_loss: 0.3587 - val_sparse_categorical_accuracy: 0.8980\n",
      "Epoch 63/200\n",
      "1/1 [==============================] - 2s 2s/step - loss: 0.3795 - sparse_categorical_accuracy: 0.8884 - val_loss: 0.3561 - val_sparse_categorical_accuracy: 0.8954\n",
      "Epoch 64/200\n",
      "1/1 [==============================] - 2s 2s/step - loss: 0.3756 - sparse_categorical_accuracy: 0.8873 - val_loss: 0.3526 - val_sparse_categorical_accuracy: 0.8997\n",
      "Epoch 65/200\n",
      "1/1 [==============================] - 2s 2s/step - loss: 0.3734 - sparse_categorical_accuracy: 0.8906 - val_loss: 0.3581 - val_sparse_categorical_accuracy: 0.8915\n",
      "Epoch 66/200\n",
      "1/1 [==============================] - 2s 2s/step - loss: 0.3762 - sparse_categorical_accuracy: 0.8835 - val_loss: 0.3687 - val_sparse_categorical_accuracy: 0.8902\n",
      "Epoch 67/200\n",
      "1/1 [==============================] - 2s 2s/step - loss: 0.3904 - sparse_categorical_accuracy: 0.8829 - val_loss: 0.4312 - val_sparse_categorical_accuracy: 0.8587\n",
      "Epoch 68/200\n",
      "1/1 [==============================] - 2s 2s/step - loss: 0.4470 - sparse_categorical_accuracy: 0.8498 - val_loss: 0.5534 - val_sparse_categorical_accuracy: 0.8044\n",
      "Epoch 69/200\n",
      "1/1 [==============================] - 2s 2s/step - loss: 0.5745 - sparse_categorical_accuracy: 0.7966 - val_loss: 0.8486 - val_sparse_categorical_accuracy: 0.7387\n",
      "Epoch 70/200\n",
      "1/1 [==============================] - 2s 2s/step - loss: 0.8618 - sparse_categorical_accuracy: 0.7351 - val_loss: 0.8455 - val_sparse_categorical_accuracy: 0.7509\n",
      "Epoch 71/200\n",
      "1/1 [==============================] - 2s 2s/step - loss: 0.8635 - sparse_categorical_accuracy: 0.7474 - val_loss: 0.4267 - val_sparse_categorical_accuracy: 0.8745\n",
      "Epoch 72/200\n",
      "1/1 [==============================] - 2s 2s/step - loss: 0.4436 - sparse_categorical_accuracy: 0.8658 - val_loss: 0.3634 - val_sparse_categorical_accuracy: 0.9055\n",
      "Epoch 73/200\n",
      "1/1 [==============================] - 2s 2s/step - loss: 0.3854 - sparse_categorical_accuracy: 0.8975 - val_loss: 0.3447 - val_sparse_categorical_accuracy: 0.9050\n",
      "Epoch 74/200\n",
      "1/1 [==============================] - 2s 2s/step - loss: 0.3646 - sparse_categorical_accuracy: 0.8969 - val_loss: 0.3333 - val_sparse_categorical_accuracy: 0.9103\n",
      "Epoch 75/200\n",
      "1/1 [==============================] - 2s 2s/step - loss: 0.3538 - sparse_categorical_accuracy: 0.9018 - val_loss: 0.3265 - val_sparse_categorical_accuracy: 0.9082\n",
      "Epoch 76/200\n",
      "1/1 [==============================] - 2s 2s/step - loss: 0.3461 - sparse_categorical_accuracy: 0.9004 - val_loss: 0.3205 - val_sparse_categorical_accuracy: 0.9120\n",
      "Epoch 77/200\n",
      "1/1 [==============================] - 2s 2s/step - loss: 0.3402 - sparse_categorical_accuracy: 0.9031 - val_loss: 0.3161 - val_sparse_categorical_accuracy: 0.9118\n",
      "Epoch 78/200\n",
      "1/1 [==============================] - 2s 2s/step - loss: 0.3353 - sparse_categorical_accuracy: 0.9025 - val_loss: 0.3119 - val_sparse_categorical_accuracy: 0.9129\n",
      "Epoch 79/200\n"
     ]
    },
    {
     "ename": "KeyboardInterrupt",
     "evalue": "",
     "output_type": "error",
     "traceback": [
      "\u001b[1;31m---------------------------------------------------------------------------\u001b[0m",
      "\u001b[1;31mKeyboardInterrupt\u001b[0m                         Traceback (most recent call last)",
      "Cell \u001b[1;32mIn[27], line 32\u001b[0m\n\u001b[0;32m     25\u001b[0m keras_model\u001b[38;5;241m.\u001b[39mcompile(\n\u001b[0;32m     26\u001b[0m   optimizer\u001b[38;5;241m=\u001b[39mtf\u001b[38;5;241m.\u001b[39mkeras\u001b[38;5;241m.\u001b[39moptimizers\u001b[38;5;241m.\u001b[39mSGD(\u001b[38;5;241m0.2\u001b[39m),\n\u001b[0;32m     27\u001b[0m   loss\u001b[38;5;241m=\u001b[39mtf\u001b[38;5;241m.\u001b[39mkeras\u001b[38;5;241m.\u001b[39mlosses\u001b[38;5;241m.\u001b[39mSparseCategoricalCrossentropy(),\n\u001b[0;32m     28\u001b[0m   metrics\u001b[38;5;241m=\u001b[39m[tf\u001b[38;5;241m.\u001b[39mkeras\u001b[38;5;241m.\u001b[39mmetrics\u001b[38;5;241m.\u001b[39mSparseCategoricalAccuracy()],\n\u001b[0;32m     29\u001b[0m )\n\u001b[0;32m     31\u001b[0m \u001b[38;5;66;03m# Train loop\u001b[39;00m\n\u001b[1;32m---> 32\u001b[0m history \u001b[38;5;241m=\u001b[39m \u001b[43mkeras_model\u001b[49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mfit\u001b[49m\u001b[43m(\u001b[49m\n\u001b[0;32m     33\u001b[0m \u001b[43m  \u001b[49m\u001b[43mtrain_images\u001b[49m\u001b[43m,\u001b[49m\n\u001b[0;32m     34\u001b[0m \u001b[43m  \u001b[49m\u001b[43mtrain_labels0\u001b[49m\u001b[43m,\u001b[49m\n\u001b[0;32m     35\u001b[0m \u001b[43m  \u001b[49m\u001b[43mbatch_size\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[38;5;28;43mlen\u001b[39;49m\u001b[43m(\u001b[49m\u001b[43mtrain_images\u001b[49m\u001b[43m)\u001b[49m\u001b[43m,\u001b[49m\n\u001b[0;32m     36\u001b[0m \u001b[43m  \u001b[49m\u001b[43mepochs\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[38;5;241;43m200\u001b[39;49m\u001b[43m,\u001b[49m\n\u001b[0;32m     37\u001b[0m \u001b[43m  \u001b[49m\u001b[43mvalidation_data\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[43m(\u001b[49m\u001b[43mtest_images\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43mtest_labels0\u001b[49m\u001b[43m)\u001b[49m\u001b[43m,\u001b[49m\n\u001b[0;32m     38\u001b[0m \u001b[43m)\u001b[49m\n",
      "File \u001b[1;32mf:\\KTU\\Neuroninių tinklų metodai\\venv\\Lib\\site-packages\\keras\\src\\utils\\traceback_utils.py:65\u001b[0m, in \u001b[0;36mfilter_traceback.<locals>.error_handler\u001b[1;34m(*args, **kwargs)\u001b[0m\n\u001b[0;32m     63\u001b[0m filtered_tb \u001b[38;5;241m=\u001b[39m \u001b[38;5;28;01mNone\u001b[39;00m\n\u001b[0;32m     64\u001b[0m \u001b[38;5;28;01mtry\u001b[39;00m:\n\u001b[1;32m---> 65\u001b[0m     \u001b[38;5;28;01mreturn\u001b[39;00m \u001b[43mfn\u001b[49m\u001b[43m(\u001b[49m\u001b[38;5;241;43m*\u001b[39;49m\u001b[43margs\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[38;5;241;43m*\u001b[39;49m\u001b[38;5;241;43m*\u001b[39;49m\u001b[43mkwargs\u001b[49m\u001b[43m)\u001b[49m\n\u001b[0;32m     66\u001b[0m \u001b[38;5;28;01mexcept\u001b[39;00m \u001b[38;5;167;01mException\u001b[39;00m \u001b[38;5;28;01mas\u001b[39;00m e:\n\u001b[0;32m     67\u001b[0m     filtered_tb \u001b[38;5;241m=\u001b[39m _process_traceback_frames(e\u001b[38;5;241m.\u001b[39m__traceback__)\n",
      "File \u001b[1;32mf:\\KTU\\Neuroninių tinklų metodai\\venv\\Lib\\site-packages\\keras\\src\\engine\\training.py:1807\u001b[0m, in \u001b[0;36mModel.fit\u001b[1;34m(self, x, y, batch_size, epochs, verbose, callbacks, validation_split, validation_data, shuffle, class_weight, sample_weight, initial_epoch, steps_per_epoch, validation_steps, validation_batch_size, validation_freq, max_queue_size, workers, use_multiprocessing)\u001b[0m\n\u001b[0;32m   1799\u001b[0m \u001b[38;5;28;01mwith\u001b[39;00m tf\u001b[38;5;241m.\u001b[39mprofiler\u001b[38;5;241m.\u001b[39mexperimental\u001b[38;5;241m.\u001b[39mTrace(\n\u001b[0;32m   1800\u001b[0m     \u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mtrain\u001b[39m\u001b[38;5;124m\"\u001b[39m,\n\u001b[0;32m   1801\u001b[0m     epoch_num\u001b[38;5;241m=\u001b[39mepoch,\n\u001b[1;32m   (...)\u001b[0m\n\u001b[0;32m   1804\u001b[0m     _r\u001b[38;5;241m=\u001b[39m\u001b[38;5;241m1\u001b[39m,\n\u001b[0;32m   1805\u001b[0m ):\n\u001b[0;32m   1806\u001b[0m     callbacks\u001b[38;5;241m.\u001b[39mon_train_batch_begin(step)\n\u001b[1;32m-> 1807\u001b[0m     tmp_logs \u001b[38;5;241m=\u001b[39m \u001b[38;5;28;43mself\u001b[39;49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mtrain_function\u001b[49m\u001b[43m(\u001b[49m\u001b[43miterator\u001b[49m\u001b[43m)\u001b[49m\n\u001b[0;32m   1808\u001b[0m     \u001b[38;5;28;01mif\u001b[39;00m data_handler\u001b[38;5;241m.\u001b[39mshould_sync:\n\u001b[0;32m   1809\u001b[0m         context\u001b[38;5;241m.\u001b[39masync_wait()\n",
      "File \u001b[1;32mf:\\KTU\\Neuroninių tinklų metodai\\venv\\Lib\\site-packages\\tensorflow\\python\\util\\traceback_utils.py:150\u001b[0m, in \u001b[0;36mfilter_traceback.<locals>.error_handler\u001b[1;34m(*args, **kwargs)\u001b[0m\n\u001b[0;32m    148\u001b[0m filtered_tb \u001b[38;5;241m=\u001b[39m \u001b[38;5;28;01mNone\u001b[39;00m\n\u001b[0;32m    149\u001b[0m \u001b[38;5;28;01mtry\u001b[39;00m:\n\u001b[1;32m--> 150\u001b[0m   \u001b[38;5;28;01mreturn\u001b[39;00m \u001b[43mfn\u001b[49m\u001b[43m(\u001b[49m\u001b[38;5;241;43m*\u001b[39;49m\u001b[43margs\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[38;5;241;43m*\u001b[39;49m\u001b[38;5;241;43m*\u001b[39;49m\u001b[43mkwargs\u001b[49m\u001b[43m)\u001b[49m\n\u001b[0;32m    151\u001b[0m \u001b[38;5;28;01mexcept\u001b[39;00m \u001b[38;5;167;01mException\u001b[39;00m \u001b[38;5;28;01mas\u001b[39;00m e:\n\u001b[0;32m    152\u001b[0m   filtered_tb \u001b[38;5;241m=\u001b[39m _process_traceback_frames(e\u001b[38;5;241m.\u001b[39m__traceback__)\n",
      "File \u001b[1;32mf:\\KTU\\Neuroninių tinklų metodai\\venv\\Lib\\site-packages\\tensorflow\\python\\eager\\polymorphic_function\\polymorphic_function.py:832\u001b[0m, in \u001b[0;36mFunction.__call__\u001b[1;34m(self, *args, **kwds)\u001b[0m\n\u001b[0;32m    829\u001b[0m compiler \u001b[38;5;241m=\u001b[39m \u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mxla\u001b[39m\u001b[38;5;124m\"\u001b[39m \u001b[38;5;28;01mif\u001b[39;00m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39m_jit_compile \u001b[38;5;28;01melse\u001b[39;00m \u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mnonXla\u001b[39m\u001b[38;5;124m\"\u001b[39m\n\u001b[0;32m    831\u001b[0m \u001b[38;5;28;01mwith\u001b[39;00m OptionalXlaContext(\u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39m_jit_compile):\n\u001b[1;32m--> 832\u001b[0m   result \u001b[38;5;241m=\u001b[39m \u001b[38;5;28;43mself\u001b[39;49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43m_call\u001b[49m\u001b[43m(\u001b[49m\u001b[38;5;241;43m*\u001b[39;49m\u001b[43margs\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[38;5;241;43m*\u001b[39;49m\u001b[38;5;241;43m*\u001b[39;49m\u001b[43mkwds\u001b[49m\u001b[43m)\u001b[49m\n\u001b[0;32m    834\u001b[0m new_tracing_count \u001b[38;5;241m=\u001b[39m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39mexperimental_get_tracing_count()\n\u001b[0;32m    835\u001b[0m without_tracing \u001b[38;5;241m=\u001b[39m (tracing_count \u001b[38;5;241m==\u001b[39m new_tracing_count)\n",
      "File \u001b[1;32mf:\\KTU\\Neuroninių tinklų metodai\\venv\\Lib\\site-packages\\tensorflow\\python\\eager\\polymorphic_function\\polymorphic_function.py:868\u001b[0m, in \u001b[0;36mFunction._call\u001b[1;34m(self, *args, **kwds)\u001b[0m\n\u001b[0;32m    865\u001b[0m   \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39m_lock\u001b[38;5;241m.\u001b[39mrelease()\n\u001b[0;32m    866\u001b[0m   \u001b[38;5;66;03m# In this case we have created variables on the first call, so we run the\u001b[39;00m\n\u001b[0;32m    867\u001b[0m   \u001b[38;5;66;03m# defunned version which is guaranteed to never create variables.\u001b[39;00m\n\u001b[1;32m--> 868\u001b[0m   \u001b[38;5;28;01mreturn\u001b[39;00m \u001b[43mtracing_compilation\u001b[49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mcall_function\u001b[49m\u001b[43m(\u001b[49m\n\u001b[0;32m    869\u001b[0m \u001b[43m      \u001b[49m\u001b[43margs\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43mkwds\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[38;5;28;43mself\u001b[39;49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43m_no_variable_creation_config\u001b[49m\n\u001b[0;32m    870\u001b[0m \u001b[43m  \u001b[49m\u001b[43m)\u001b[49m\n\u001b[0;32m    871\u001b[0m \u001b[38;5;28;01melif\u001b[39;00m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39m_variable_creation_config \u001b[38;5;129;01mis\u001b[39;00m \u001b[38;5;129;01mnot\u001b[39;00m \u001b[38;5;28;01mNone\u001b[39;00m:\n\u001b[0;32m    872\u001b[0m   \u001b[38;5;66;03m# Release the lock early so that multiple threads can perform the call\u001b[39;00m\n\u001b[0;32m    873\u001b[0m   \u001b[38;5;66;03m# in parallel.\u001b[39;00m\n\u001b[0;32m    874\u001b[0m   \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39m_lock\u001b[38;5;241m.\u001b[39mrelease()\n",
      "File \u001b[1;32mf:\\KTU\\Neuroninių tinklų metodai\\venv\\Lib\\site-packages\\tensorflow\\python\\eager\\polymorphic_function\\tracing_compilation.py:139\u001b[0m, in \u001b[0;36mcall_function\u001b[1;34m(args, kwargs, tracing_options)\u001b[0m\n\u001b[0;32m    137\u001b[0m bound_args \u001b[38;5;241m=\u001b[39m function\u001b[38;5;241m.\u001b[39mfunction_type\u001b[38;5;241m.\u001b[39mbind(\u001b[38;5;241m*\u001b[39margs, \u001b[38;5;241m*\u001b[39m\u001b[38;5;241m*\u001b[39mkwargs)\n\u001b[0;32m    138\u001b[0m flat_inputs \u001b[38;5;241m=\u001b[39m function\u001b[38;5;241m.\u001b[39mfunction_type\u001b[38;5;241m.\u001b[39munpack_inputs(bound_args)\n\u001b[1;32m--> 139\u001b[0m \u001b[38;5;28;01mreturn\u001b[39;00m \u001b[43mfunction\u001b[49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43m_call_flat\u001b[49m\u001b[43m(\u001b[49m\u001b[43m  \u001b[49m\u001b[38;5;66;43;03m# pylint: disable=protected-access\u001b[39;49;00m\n\u001b[0;32m    140\u001b[0m \u001b[43m    \u001b[49m\u001b[43mflat_inputs\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43mcaptured_inputs\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[43mfunction\u001b[49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mcaptured_inputs\u001b[49m\n\u001b[0;32m    141\u001b[0m \u001b[43m\u001b[49m\u001b[43m)\u001b[49m\n",
      "File \u001b[1;32mf:\\KTU\\Neuroninių tinklų metodai\\venv\\Lib\\site-packages\\tensorflow\\python\\eager\\polymorphic_function\\concrete_function.py:1323\u001b[0m, in \u001b[0;36mConcreteFunction._call_flat\u001b[1;34m(self, tensor_inputs, captured_inputs)\u001b[0m\n\u001b[0;32m   1319\u001b[0m possible_gradient_type \u001b[38;5;241m=\u001b[39m gradients_util\u001b[38;5;241m.\u001b[39mPossibleTapeGradientTypes(args)\n\u001b[0;32m   1320\u001b[0m \u001b[38;5;28;01mif\u001b[39;00m (possible_gradient_type \u001b[38;5;241m==\u001b[39m gradients_util\u001b[38;5;241m.\u001b[39mPOSSIBLE_GRADIENT_TYPES_NONE\n\u001b[0;32m   1321\u001b[0m     \u001b[38;5;129;01mand\u001b[39;00m executing_eagerly):\n\u001b[0;32m   1322\u001b[0m   \u001b[38;5;66;03m# No tape is watching; skip to running the function.\u001b[39;00m\n\u001b[1;32m-> 1323\u001b[0m   \u001b[38;5;28;01mreturn\u001b[39;00m \u001b[38;5;28;43mself\u001b[39;49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43m_inference_function\u001b[49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mcall_preflattened\u001b[49m\u001b[43m(\u001b[49m\u001b[43margs\u001b[49m\u001b[43m)\u001b[49m\n\u001b[0;32m   1324\u001b[0m forward_backward \u001b[38;5;241m=\u001b[39m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39m_select_forward_and_backward_functions(\n\u001b[0;32m   1325\u001b[0m     args,\n\u001b[0;32m   1326\u001b[0m     possible_gradient_type,\n\u001b[0;32m   1327\u001b[0m     executing_eagerly)\n\u001b[0;32m   1328\u001b[0m forward_function, args_with_tangents \u001b[38;5;241m=\u001b[39m forward_backward\u001b[38;5;241m.\u001b[39mforward()\n",
      "File \u001b[1;32mf:\\KTU\\Neuroninių tinklų metodai\\venv\\Lib\\site-packages\\tensorflow\\python\\eager\\polymorphic_function\\atomic_function.py:216\u001b[0m, in \u001b[0;36mAtomicFunction.call_preflattened\u001b[1;34m(self, args)\u001b[0m\n\u001b[0;32m    214\u001b[0m \u001b[38;5;28;01mdef\u001b[39;00m \u001b[38;5;21mcall_preflattened\u001b[39m(\u001b[38;5;28mself\u001b[39m, args: Sequence[core\u001b[38;5;241m.\u001b[39mTensor]) \u001b[38;5;241m-\u001b[39m\u001b[38;5;241m>\u001b[39m Any:\n\u001b[0;32m    215\u001b[0m \u001b[38;5;250m  \u001b[39m\u001b[38;5;124;03m\"\"\"Calls with flattened tensor inputs and returns the structured output.\"\"\"\u001b[39;00m\n\u001b[1;32m--> 216\u001b[0m   flat_outputs \u001b[38;5;241m=\u001b[39m \u001b[38;5;28;43mself\u001b[39;49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mcall_flat\u001b[49m\u001b[43m(\u001b[49m\u001b[38;5;241;43m*\u001b[39;49m\u001b[43margs\u001b[49m\u001b[43m)\u001b[49m\n\u001b[0;32m    217\u001b[0m   \u001b[38;5;28;01mreturn\u001b[39;00m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39mfunction_type\u001b[38;5;241m.\u001b[39mpack_output(flat_outputs)\n",
      "File \u001b[1;32mf:\\KTU\\Neuroninių tinklų metodai\\venv\\Lib\\site-packages\\tensorflow\\python\\eager\\polymorphic_function\\atomic_function.py:251\u001b[0m, in \u001b[0;36mAtomicFunction.call_flat\u001b[1;34m(self, *args)\u001b[0m\n\u001b[0;32m    249\u001b[0m \u001b[38;5;28;01mwith\u001b[39;00m record\u001b[38;5;241m.\u001b[39mstop_recording():\n\u001b[0;32m    250\u001b[0m   \u001b[38;5;28;01mif\u001b[39;00m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39m_bound_context\u001b[38;5;241m.\u001b[39mexecuting_eagerly():\n\u001b[1;32m--> 251\u001b[0m     outputs \u001b[38;5;241m=\u001b[39m \u001b[38;5;28;43mself\u001b[39;49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43m_bound_context\u001b[49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mcall_function\u001b[49m\u001b[43m(\u001b[49m\n\u001b[0;32m    252\u001b[0m \u001b[43m        \u001b[49m\u001b[38;5;28;43mself\u001b[39;49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mname\u001b[49m\u001b[43m,\u001b[49m\n\u001b[0;32m    253\u001b[0m \u001b[43m        \u001b[49m\u001b[38;5;28;43mlist\u001b[39;49m\u001b[43m(\u001b[49m\u001b[43margs\u001b[49m\u001b[43m)\u001b[49m\u001b[43m,\u001b[49m\n\u001b[0;32m    254\u001b[0m \u001b[43m        \u001b[49m\u001b[38;5;28;43mlen\u001b[39;49m\u001b[43m(\u001b[49m\u001b[38;5;28;43mself\u001b[39;49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mfunction_type\u001b[49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mflat_outputs\u001b[49m\u001b[43m)\u001b[49m\u001b[43m,\u001b[49m\n\u001b[0;32m    255\u001b[0m \u001b[43m    \u001b[49m\u001b[43m)\u001b[49m\n\u001b[0;32m    256\u001b[0m   \u001b[38;5;28;01melse\u001b[39;00m:\n\u001b[0;32m    257\u001b[0m     outputs \u001b[38;5;241m=\u001b[39m make_call_op_in_graph(\n\u001b[0;32m    258\u001b[0m         \u001b[38;5;28mself\u001b[39m,\n\u001b[0;32m    259\u001b[0m         \u001b[38;5;28mlist\u001b[39m(args),\n\u001b[0;32m    260\u001b[0m         \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39m_bound_context\u001b[38;5;241m.\u001b[39mfunction_call_options\u001b[38;5;241m.\u001b[39mas_attrs(),\n\u001b[0;32m    261\u001b[0m     )\n",
      "File \u001b[1;32mf:\\KTU\\Neuroninių tinklų metodai\\venv\\Lib\\site-packages\\tensorflow\\python\\eager\\context.py:1486\u001b[0m, in \u001b[0;36mContext.call_function\u001b[1;34m(self, name, tensor_inputs, num_outputs)\u001b[0m\n\u001b[0;32m   1484\u001b[0m cancellation_context \u001b[38;5;241m=\u001b[39m cancellation\u001b[38;5;241m.\u001b[39mcontext()\n\u001b[0;32m   1485\u001b[0m \u001b[38;5;28;01mif\u001b[39;00m cancellation_context \u001b[38;5;129;01mis\u001b[39;00m \u001b[38;5;28;01mNone\u001b[39;00m:\n\u001b[1;32m-> 1486\u001b[0m   outputs \u001b[38;5;241m=\u001b[39m \u001b[43mexecute\u001b[49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mexecute\u001b[49m\u001b[43m(\u001b[49m\n\u001b[0;32m   1487\u001b[0m \u001b[43m      \u001b[49m\u001b[43mname\u001b[49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mdecode\u001b[49m\u001b[43m(\u001b[49m\u001b[38;5;124;43m\"\u001b[39;49m\u001b[38;5;124;43mutf-8\u001b[39;49m\u001b[38;5;124;43m\"\u001b[39;49m\u001b[43m)\u001b[49m\u001b[43m,\u001b[49m\n\u001b[0;32m   1488\u001b[0m \u001b[43m      \u001b[49m\u001b[43mnum_outputs\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[43mnum_outputs\u001b[49m\u001b[43m,\u001b[49m\n\u001b[0;32m   1489\u001b[0m \u001b[43m      \u001b[49m\u001b[43minputs\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[43mtensor_inputs\u001b[49m\u001b[43m,\u001b[49m\n\u001b[0;32m   1490\u001b[0m \u001b[43m      \u001b[49m\u001b[43mattrs\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[43mattrs\u001b[49m\u001b[43m,\u001b[49m\n\u001b[0;32m   1491\u001b[0m \u001b[43m      \u001b[49m\u001b[43mctx\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[38;5;28;43mself\u001b[39;49m\u001b[43m,\u001b[49m\n\u001b[0;32m   1492\u001b[0m \u001b[43m  \u001b[49m\u001b[43m)\u001b[49m\n\u001b[0;32m   1493\u001b[0m \u001b[38;5;28;01melse\u001b[39;00m:\n\u001b[0;32m   1494\u001b[0m   outputs \u001b[38;5;241m=\u001b[39m execute\u001b[38;5;241m.\u001b[39mexecute_with_cancellation(\n\u001b[0;32m   1495\u001b[0m       name\u001b[38;5;241m.\u001b[39mdecode(\u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mutf-8\u001b[39m\u001b[38;5;124m\"\u001b[39m),\n\u001b[0;32m   1496\u001b[0m       num_outputs\u001b[38;5;241m=\u001b[39mnum_outputs,\n\u001b[1;32m   (...)\u001b[0m\n\u001b[0;32m   1500\u001b[0m       cancellation_manager\u001b[38;5;241m=\u001b[39mcancellation_context,\n\u001b[0;32m   1501\u001b[0m   )\n",
      "File \u001b[1;32mf:\\KTU\\Neuroninių tinklų metodai\\venv\\Lib\\site-packages\\tensorflow\\python\\eager\\execute.py:53\u001b[0m, in \u001b[0;36mquick_execute\u001b[1;34m(op_name, num_outputs, inputs, attrs, ctx, name)\u001b[0m\n\u001b[0;32m     51\u001b[0m \u001b[38;5;28;01mtry\u001b[39;00m:\n\u001b[0;32m     52\u001b[0m   ctx\u001b[38;5;241m.\u001b[39mensure_initialized()\n\u001b[1;32m---> 53\u001b[0m   tensors \u001b[38;5;241m=\u001b[39m \u001b[43mpywrap_tfe\u001b[49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mTFE_Py_Execute\u001b[49m\u001b[43m(\u001b[49m\u001b[43mctx\u001b[49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43m_handle\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43mdevice_name\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43mop_name\u001b[49m\u001b[43m,\u001b[49m\n\u001b[0;32m     54\u001b[0m \u001b[43m                                      \u001b[49m\u001b[43minputs\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43mattrs\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43mnum_outputs\u001b[49m\u001b[43m)\u001b[49m\n\u001b[0;32m     55\u001b[0m \u001b[38;5;28;01mexcept\u001b[39;00m core\u001b[38;5;241m.\u001b[39m_NotOkStatusException \u001b[38;5;28;01mas\u001b[39;00m e:\n\u001b[0;32m     56\u001b[0m   \u001b[38;5;28;01mif\u001b[39;00m name \u001b[38;5;129;01mis\u001b[39;00m \u001b[38;5;129;01mnot\u001b[39;00m \u001b[38;5;28;01mNone\u001b[39;00m:\n",
      "\u001b[1;31mKeyboardInterrupt\u001b[0m: "
     ]
    }
   ],
   "source": [
    "# train mnist\n",
    "\n",
    "mnist = tf.keras.datasets.mnist\n",
    "(train_images0, train_labels0), (test_images0, test_labels0) = mnist.load_data()\n",
    "\n",
    "test_images  = test_images0.reshape(10000, 28, 28)\n",
    "train_images = train_images0.reshape(60000, 28, 28)\n",
    "\n",
    "test_images  = test_images/255.0\n",
    "train_images = train_images/255.0\n",
    "\n",
    "keras_model = tf.keras.models.Sequential([\n",
    "  layers.Conv2D(8, (2, 2), activation='relu', input_shape=(28, 28, 1)),\n",
    "  layers.MaxPooling2D((2, 2)),\n",
    "  layers.Conv2D(16, (2, 2), activation='relu'),\n",
    "  layers.MaxPooling2D((2, 2)),\n",
    "  layers.Flatten(),\n",
    "  layers.Dense(10, activation='softmax')\n",
    "])\n",
    "\n",
    "keras_model.build(input_shape=[None,784])\n",
    "\n",
    "keras_model.summary()\n",
    "\n",
    "keras_model.compile(\n",
    "  optimizer=tf.keras.optimizers.SGD(0.2),\n",
    "  loss=tf.keras.losses.CategoricalHinge(),\n",
    "  metrics=[tf.keras.metrics.SparseCategoricalAccuracy()],\n",
    ")\n",
    "\n",
    "# Train loop\n",
    "history = keras_model.fit(\n",
    "  train_images,\n",
    "  train_labels0,\n",
    "  batch_size=len(train_images),\n",
    "  epochs=200,\n",
    "  validation_data=(test_images, test_labels0),\n",
    ")\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Key: loss\n",
      "Key: sparse_categorical_accuracy\n",
      "Key: val_loss\n",
      "Key: val_sparse_categorical_accuracy\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAngAAAHHCAYAAAAs4yUHAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjguMywgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy/H5lhTAAAACXBIWXMAAA9hAAAPYQGoP6dpAAB16ElEQVR4nO3deXhU5f3+8ffMZN8JCVlISNh3grJEUBQhiqCouIPfCqJYFSqKuNVWsa1L1SIuFNtfVShWwQW1VURZBATZERVZhBiWEJKQhOwhy8z5/THJwJAACUwyyXC/rutcM3PmnDOfhyHk5nnOc47JMAwDEREREfEYZncXICIiIiKupYAnIiIi4mEU8EREREQ8jAKeiIiIiIdRwBMRERHxMAp4IiIiIh5GAU9ERETEwyjgiYiIiHgYBTwRERERD6OAJyLSjJlMJmbMmNHg/fbt24fJZGLu3Lkur0lEmj8FPBGRM5g7dy4mkwmTycSaNWtqvW8YBvHx8ZhMJq655ho3VHj2Vq5ciclk4qOPPnJ3KSLiQgp4IiL15Ofnx3vvvVdr/apVq0hPT8fX19cNVYmI1KaAJyJST6NGjeLDDz+kqqrKaf17771Hv379iI6OdlNlIiLOFPBEROpp7Nix5ObmsnTpUse6iooKPvroI8aNG1fnPiUlJTz88MPEx8fj6+tL165defnllzEMw2m78vJyHnroISIjIwkODubaa68lPT29zmMeOnSIiRMnEhUVha+vLz179uTtt992XUPr8Ouvv3LzzTcTHh5OQEAAF110EV988UWt7V5//XV69uxJQEAArVq1on///k69nkVFRTz44IMkJibi6+tLmzZtuOKKK9i6dWuj1i9yvlHAExGpp8TERAYNGsT777/vWPfll19SUFDAbbfdVmt7wzC49tpreeWVV7jqqquYOXMmXbt25ZFHHmHatGlO2959993MmjWLK6+8khdeeAFvb2+uvvrqWsfMysrioosuYtmyZUyZMoVXX32VTp06cddddzFr1iyXt7nmMwcPHsxXX33F/fffz7PPPsuxY8e49tpr+eSTTxzb/b//9/944IEH6NGjB7NmzeKZZ56hb9++bNiwwbHNvffey5w5c7jxxhv5+9//zvTp0/H392fnzp2NUrvIecsQEZHTeueddwzA2LRpk/HGG28YwcHBRmlpqWEYhnHzzTcbl19+uWEYhpGQkGBcffXVjv0+/fRTAzD+8pe/OB3vpptuMkwmk7F3717DMAxj27ZtBmDcf//9TtuNGzfOAIynn37ase6uu+4yYmJijJycHKdtb7vtNiM0NNRRV1pamgEY77zzzmnb9s033xiA8eGHH55ymwcffNAAjG+//daxrqioyGjfvr2RmJhoWK1WwzAM47rrrjN69ux52s8LDQ01Jk+efNptROTcqQdPRKQBbrnlFsrKyvj8888pKiri888/P+Xw7OLFi7FYLDzwwANO6x9++GEMw+DLL790bAfU2u7BBx90em0YBh9//DGjR4/GMAxycnIcy4gRIygoKGiUoc7FixczcOBALrnkEse6oKAg7rnnHvbt28eOHTsACAsLIz09nU2bNp3yWGFhYWzYsIGMjAyX1ykixyngiYg0QGRkJCkpKbz33nssWrQIq9XKTTfdVOe2+/fvJzY2luDgYKf13bt3d7xf82g2m+nYsaPTdl27dnV6feTIEfLz8/nnP/9JZGSk03LnnXcCkJ2d7ZJ2ntyOk2upqx2PPfYYQUFBDBw4kM6dOzN58mTWrl3rtM+LL77I9u3biY+PZ+DAgcyYMYNff/3V5TWLnO+83F2AiEhLM27cOCZNmkRmZiYjR44kLCysST7XZrMB8H//93+MHz++zm369OnTJLXUpXv37uzevZvPP/+cJUuW8PHHH/P3v/+dp556imeeeQaw94AOGTKETz75hK+//pqXXnqJv/71ryxatIiRI0e6rXYRT6MePBGRBhozZgxms5n169efcngWICEhgYyMDIqKipzW79q1y/F+zaPNZiM1NdVpu927dzu9rplha7VaSUlJqXNp06aNK5pYqx0n11JXOwACAwO59dZbeeeddzhw4ABXX321Y1JGjZiYGO6//34+/fRT0tLSaN26Nc8++6zL6xY5nyngiYg0UFBQEHPmzGHGjBmMHj36lNuNGjUKq9XKG2+84bT+lVdewWQyOXqsah5fe+01p+1OnhVrsVi48cYb+fjjj9m+fXutzzty5MjZNOeMRo0axcaNG1m3bp1jXUlJCf/85z9JTEykR48eAOTm5jrt5+PjQ48ePTAMg8rKSqxWKwUFBU7btGnThtjYWMrLyxuldpHzlYZoRUTOwqmGSE80evRoLr/8cp588kn27dtHUlISX3/9NZ999hkPPvig45y7vn37MnbsWP7+979TUFDA4MGDWb58OXv37q11zBdeeIFvvvmG5ORkJk2aRI8ePcjLy2Pr1q0sW7aMvLy8s2rPxx9/7OiRO7mdjz/+OO+//z4jR47kgQceIDw8nHnz5pGWlsbHH3+M2WzvK7jyyiuJjo7m4osvJioqip07d/LGG29w9dVXExwcTH5+PnFxcdx0000kJSURFBTEsmXL2LRpE3/729/Oqm4ROQX3TuIVEWn+TrxMyumcfJkUw7BfTuShhx4yYmNjDW9vb6Nz587GSy+9ZNhsNqftysrKjAceeMBo3bq1ERgYaIwePdo4ePBgrcukGIZhZGVlGZMnTzbi4+MNb29vIzo62hg+fLjxz3/+07FNQy+Tcqql5tIoqampxk033WSEhYUZfn5+xsCBA43PP//c6Vj/+Mc/jEsvvdRo3bq14evra3Ts2NF45JFHjIKCAsMwDKO8vNx45JFHjKSkJCM4ONgIDAw0kpKSjL///e+nrVFEGs5kGCddTl1EREREWjSdgyciIiLiYRTwRERERDyMAp6IiIiIh1HAExEREfEwCngiIiIiHkYBT0RERMTD6ELH5ymbzUZGRgbBwcGYTCZ3lyMiIiL1YBgGRUVFxMbGOi4yXhcFvPNURkYG8fHx7i5DREREzsLBgweJi4s75fsKeOep4OBgwP4XJCQkxKXHrqys5Ouvv+bKK6/E29vbpcduDjy9feD5bfT09oHa6Ak8vX2gNp6NwsJC4uPjHb/HT0UB7zxVMywbEhLSKAEvICCAkJAQj/yB9fT2gee30dPbB2qjJ/D09oHaeC7OdHqVJlmIiIiIeBgFPBEREREPo4AnIiIi4mF0Dp6IiEgLZ7VaqaysdHcZDVZZWYmXlxfHjh3DarW6u5xG0dA2ent7Y7FYzvlzFfBERERaKMMwyMzMJD8/392lnBXDMIiOjubgwYMee03Ws2ljWFgY0dHR5/RnooAnIiLSQtWEuzZt2hAQENDiQpLNZqO4uJigoKDTXrS3JWtIGw3DoLS0lOzsbABiYmLO+nMV8ERERFogq9XqCHetW7d2dzlnxWazUVFRgZ+fn0cHvIa00d/fH4Ds7GzatGlz1sO1nvmnKSIi4uFqzrkLCAhwcyXiajXf6bmcV6mAJyIi0oK1tGFZOTNXfKcKeCIiIiIeRgFPREREWqzExERmzZrl7jKaHU2yEBERkSY1dOhQ+vbty8yZM8/5WJs2bSIwMNAFVXkWBTxxqYoqG4eOllFQ4e5KRESkpTIMA6vVipfXmWNKZGRkE1TU8miIVlzqr0t2cfnMb1l5WH+1RESktgkTJrBq1SpeffVVLBYLrVq1Yu7cuZhMJr788kv69euHr68va9asITU1leuuu46oqCiCgoIYMGAAy5YtczreyUO0JpOJf/3rX4wZM4aAgAA6d+7Mf//73yZupfvpt7C4VFSIL4B68ERE3MAwDEorqtyyGIZRrxpfffVVBg0axKRJkzh06BC7du0iPj4egMcff5wXXniBnTt30qdPH4qLixk1ahTLly/n+++/56qrrmL06NEcOHDgtJ/xzDPPcMstt/Djjz8yatQobr/9dvLy8s75z7cl0RCtuFRUiB8ABRWati8i0tTKKq30eOort3z2jj+NIMDnzLEiNDQUHx8fAgICiI6OJiAggEOHDgHwpz/9iSuuuMKxbXh4OElJSY7Xf/7zn/nkk0/473//y5QpU075GRMmTGDs2LEAPPfcc7z22mts3LiRq6666myb1+KoB09cKro64OWrB09ERBqof//+Tq+Li4uZPn063bt3JywsjKCgIHbu3HnGHrw+ffo4ngcGBhISEuK4/df5Qj144lLRoTU9eNS7u15ERFzD39vCjj+NcNtnn6uTZ8NOnz6dpUuX8vLLL9OpUyf8/f256aabqKg4fS+Ct7e302uTyYTNZjvn+loSBTxxqZoh2kqbicJjVUT4+Li5IhGR84fJZKrXMKm7+fj4YLVaz7jd2rVrmTBhAmPGjAHsPXr79u1r5Oo8g4ZoxaX8vC2E+dv/55RVeMzN1YiISHOUmJjIhg0b2LdvH7m5uafsXevcuTOLFi1i27Zt/PDDD4wbN+6864k7Wwp44nI1M2mzCsvdXImIiDRH06dPx2Kx0KtXLzp16nTKc+pmzpxJq1atGDx4MKNHj2bEiBFceOGFTVxty9T8+3GlxYkK8WV3VjFZRQp4IiJSW5cuXVi3bh02m43CwkJCQkKYOHFire0SExNZsWKF07rJkyc7vT55yLau87/z8/PPueaWRj144nJtgu3n4akHT0RExD0U8MTljg/R6hw8ERERd1DAE5fTOXgiIiLupYAnLldzqZSsIvXgiYiIuIMCnrhcVLB68ERERNxJAU9cLrp6iDa3pIJKq65XJCIi0tQU8MTlWgX4YDEZGAZk61IpIiIiTU4BT1zObDYRWn2HsswCnYcnIiLS1BTwpFGEVN/nOVuXShEREWlyCnjSKMJ87FcSz1TAExERF0tMTGTWrFmO1yaTiU8//fSU2+/btw+TycS2bdvO6XNddZymoFuVSaNwDNEq4ImISCM7fPgwrVq1cukxJ0yYQH5+vlNwjI+P5/Dhw0RERLj0sxqDAp40itDqHrwsnYMnIiKNLDo6ukk+x2KxNNlnnSsN0UqjUA+eiIjU5Z///CexsbHYbM6X0bruuuuYOHEiqampXHfddURFRREUFMSAAQNYtmzZaY958hDtxo0bueCCC/Dz86N///58//33TttbrVbuuusu2rdvj7+/P127duXVV191vD9jxgzmzZvHZ599hslkwmQysXLlyjqHaFetWsXAgQPx9fUlJiaGxx9/nKqqKsf7w4YN47HHHuOxxx4jPDyc6OhoZsyY0fA/uAZSD540Csc5eOrBExFpOoYBlaXu+WzvADCZzrjZzTffzO9+9zu++eYbLr/8cgDy8vJYsmQJixcvpri4mFGjRvHss8/i6+vLv//9b0aPHs3u3btp167dGY9fXFzMNddcwxVXXMG7775LWloaU6dOddrGZrMRFxfHhx9+SOvWrfnuu++45557iImJ4ZZbbmH69Ons3LmTwsJC3nnnHQDCw8PJyMhwOs6hQ4cYNWoUEyZM4N///je7du1i0qRJ+Pn5OYW4999/n2nTprFhwwbWrVvHhAkTuPjii7niiivO2J6zpYAnjSLMfq1jDhccwzAMTPX4oRcRkXNUWQrPxbrns3+fAT6BZ9ysVatWjBw5kvfee88R8D766CMiIiK4/PLLMZvNJCUlObb/85//zCeffMJ///tfpkyZcsbjv/fee9hsNt566y38/Pzo2bMn6enp3HfffY5tvL29eeaZZxyv27dvz7p16/jggw+45ZZbCAoKwt/fn/Ly8tMOyf79738nPj6eN954A5PJRLdu3cjIyOCxxx7jqaeewmy2D5T27NnT8bpz58688cYbLF++vFEDnoZopVGE+dj/I1deZSOvpMLd5YiISDNy++238/HHH1Nebr8Y/vvvv89tt92G2WymuLiY6dOn0717d8LCwggKCmLnzp0cOHCgXsfeuXMnffr0wc/Pz7Fu0KBBtbabPXs2/fr1IzIykqCgIP75z3/W+zNO/KxBgwY5dWJcfPHFFBcXk56e7ljXs2dPp/1iYmLIzs5u0Gc1lHrwpFF4mSEyyJfsonIy8o/ROsjX3SWJiHg+7wB7T5q7PrueRo8ejWEYfPHFF3Tr1o1vv/2WV155BYDp06ezdOlSXn75ZTp16oS/vz833XQTFRWu6yxYsGAB06dP529/+xuDBg0iODiYl156iQ0bNrjsM07k7e3t9NpkMtU6B9HVFPCk0USHVge8gjJ6x4W6uxwREc9nMtVrmNTd/Pz8uOGGG3jvvfdISkqia9euXHjhhQCsXbuWCRMmMGbMGMB+Tt2+ffvqfezu3bszf/58jh075ujFW79+vdM2a9euZfDgwdx///2OdampqU7b+Pj4YLVaz/hZH3/8sdOpSGvXriU4OJi4uLh619wYNEQrjSY21B+AjPwyN1ciIiLNze23387ixYv5z3/+w7hx4xzrO3fuzKJFi9i2bRs//PAD48aNa1Bv17hx4zCZTEyaNIkdO3awePFiXn75ZadtOnfuzObNm/nqq6/45Zdf+OMf/8imTZuctklMTOTHH39k9+7d5OTkUFlZWeuz7r//fg4ePMjvfvc7du3axWeffcbTTz/NtGnTHOffuYsCnjSa2FD7/5wU8ERE5GTDhg0jPDycPXv2MHbsWMf6mTNn0qpVKwYPHszo0aMZMWKEo3evPoKCgvjf//7HTz/9xAUXXMCTTz7JX//6V6dtfvvb33LDDTdw6623kpycTG5urlNvHsCkSZPo2rUr/fv3JzIykrVr19b6rLZt27J48WI2btxIUlIS9957L3fddRd/+MMfGvin4XoaopVGE10T8HSpFBEROYnZbCY9PZ3CwkJCQkIc6xMTE1mxYoXTtpMnT3Z6ffKQrWEYTq8vuuiiWrcTO3EbX19f3nnnHcclUGo8//zzjueRkZF8/fXXteo++bMuu+wyNm7cWGu7GitWrKCwsNBp3eluq+Yq6sGTRqMePBEREfdQwJNGE6OAJyIi4hYKeNJoYsPsAS+7qJxKa+NOBxcREZHjFPCk0YQH+ODjZcYwdMsyERGRpqSAJ43GbDY5hmkPK+CJiDSKk0/6l5bPFd+pAp6HGDNmDK1ateKmm25ydylOdC08EZHGUXN3hNLSUjdXIq5W852efAeMhtBlUjzE1KlTmThxIvPmzXN3KU5iw+wB75ACnoiIS1ksFsLCwhz3NA0ICHC6J2pLYLPZqKio4NixY26/MHBjaUgbDcOgtLSU7OxswsLCsFgsZ/25CngeYujQoaxcudLdZdRSM9HicIECnoiIq0VHRwM0+o3rG4thGJSVleHv79/iwml9nU0bw8LCHN/t2XJ7wFu9ejUvvfQSW7Zs4fDhw3zyySdcf/31Ltln9uzZvPTSS2RmZpKUlMTrr7/OwIEDAZgxYwbPPPOM0/Zdu3Zl165drmqay2ptyWp68DLydQ6eiIirmUwmYmJiaNOmTZ230mruKisrWb16NZdeeuk5DUc2Zw1to7e39zn13NVwe8ArKSkhKSmJiRMncsMNN7hsn4ULFzJt2jTefPNNkpOTmTVrFiNGjGD37t20adMGgJ49e7Js2TLHPl5ep//jWLt2LQMHDqz1Be3YsYPWrVsTFRXVaLX27duXqqqqWvt+/fXXxMbGnrZudzoe8NSDJyLSWCwWi0tCQVOzWCxUVVXh5+fnsQHPXW10e8AbOXIkI0eOdPk+M2fOZNKkSdx5550AvPnmm3zxxRe8/fbbPP7444A90NW3C9RmszF58mQ6d+7MggULHD9Iu3fvZtiwYUybNo1HH3200Wo9+ZYrLYXuZiEiItL0PPKMxoqKCrZs2UJKSopjndlsJiUlhXXr1jnW7dmzh9jYWDp06MDtt9/OgQMHTnlMs9nM4sWL+f7777njjjuw2WykpqYybNgwrr/++jrDnStrdZXZs2fTo0cPBgwY4PJj1yWmugev8FgVRcda3vCBiIhIS+SRAS8nJwer1VpryDQqKorMzEwAkpOTmTt3LkuWLGHOnDmkpaUxZMgQioqKTnnc2NhYVqxYwZo1axg3bhzDhg0jJSWFOXPmNGqt9ZGSksLNN9/M4sWLiYuLO2U4nDx5Mjt27GDTpk1nXXNDBPl6ERZg75JOP6pePBERkabg9iFadzlx2LRPnz4kJyeTkJDABx98wF133XXK/dq1a8f8+fO57LLL6NChA2+99VazmPlz4rmEzU1CeAD5pQXszy2le0yIu8sRERHxeB7ZgxcREYHFYiErK8tpfVZW1inPuQsLC6NLly7s3bv3tMfOysrinnvuYfTo0ZSWlvLQQw81ea0tTXx4AAAH83QxThERkabgkQHPx8eHfv36sXz5csc6m83G8uXLGTRoUJ37FBcXk5qaSkxMzCmPm5OTw/Dhw+nevTuLFi1i+fLlLFy4kOnTpzdprS1NQmt7wNufV+LmSkRERM4Pbh+iLS4uduo1S0tLY9u2bYSHh9OuXTveeOMNPvnkE6cAdKZ9AKZNm8b48ePp378/AwcOZNasWZSUlDhmqk6fPp3Ro0eTkJBARkYGTz/9NBaLhbFjx9ZZp81mY+TIkSQkJLBw4UK8vLzo0aMHS5cuZdiwYbRt27bO3jxX1NrSJYQHArA/Vz14IiIiTcHtAW/z5s1cfvnljtfTpk0DYPz48cydO5ecnBxSU1MbtA/ArbfeypEjR3jqqafIzMykb9++LFmyxDGZIT09nbFjx5Kbm0tkZCSXXHIJ69evJzIyss46zWYzzz33HEOGDMHHx8exPikpiWXLlp1yP1fU2tK1q+7BO6AhWhERkSbh9oA3dOhQDMM45fszZsxgxowZDdqnxpQpU5gyZUqd7y1YsKBBdQJcccUVda6/4IILTrmPK2pt6WqGaA8dLaPKasPL4pFnBoiIiDQb+k0rjS4q2A8fLzNVNkO3LBMREWkCCnjS6MxmE/Gt7Bc81kQLERGRxqeAJ00iobV9ooXOwxMREWl8CnjSJNpVXwvvgGbSioiINDoFPGkSjmvhKeCJiIg0OgU8aRLHL3asgCciItLYFPCkSRwfoi2p12VjRERE5Owp4EmTiGsVgMkEJRVWcksq3F2OiIiIR1PAkybh520hOsQP0ExaERGRxqaAJ02mZph2f66uhSciItKYFPCkyXSIDAIgNVsBT0REpDEp4EmT6dzGHvD2ZBe5uRIRERHPpoAnTaZzVE3AK3ZzJSIiIp5NAU+aTJeoYMB+sePyKqubqxEREfFcCnjSZNoE+xLs54XVZrAvRzNpRUREGosCnjQZk8mk8/BERESagAKeNKnObezDtHuydB6eiIhIY1HAkyZVM9FiryZaiIiINBoFPGlSnTREKyIi0ugU8KRJda6eSZuWU0Kl1ebmakRERDyTAp40qdhQPwJ9LFRaDfbnaiatiIhIY1DAkyZlMpkcw7R7NUwrIiLSKBTwpMl10kxaERGRRqWAJ02uZibt7iz14ImIiDQGBTxpcj1iQgD4OaPQzZWIiIh4JgU8aXK924YC9pm0hccq3VyNiIiI51HAkybXKtCHtmH+AGw/VODmakRERDyPAp64RZ84ey+eAp6IiIjrKeCJW/SqHqb96ZDOwxMREXE1BTxxi5rz8NSDJyIi4noKeOIWmmghIiLSeBTwxC1OnGjxs4ZpRUREXEoBT9xGw7QiIiKNQwFP3KZ3XM1ECwU8ERERV1LAE7c5PpNWAU9ERMSVFPDEbfrGhWEy2SdaHCkqd3c5IiIiHkMBT9wmNMCbrlHBAGzal+fmakRERDyHAp641UUdWgOw4ddcN1ciIiLiORTwxK2S24cDsCFNPXgiIiKuooAnbjWgOuDtyiziaEmFm6sRERHxDAp44lYRQb50ahME6Dw8ERERV1HAE7cbqGFaERERl1LAE7c7fh6eJlqIiIi4ggKeuF1ye/tM2h0ZhRQeq3RzNSIiIi2fAp64XXSoH+0jArEZ8N3eHHeXIyIi0uIp4EmzcHnXNgAs35nt5kpERERaPgU8aRaGd7cHvG92H8FmM9xcjYiISMumgCfNwoDEcIJ8vcgpLuenQwXuLkdERKRFU8CTZsHHy8yQzhEALN+lYVoREZFzoYAnzcawbvZh2hW7stxciYiISMumgCfNxtCubTCZYPuhQrIKj7m7HBERkRZLAU+ajchgX/rEhQGwbKd68URERM6WAp40K1f1jAbg8x8Ou7kSERGRlksBT5qV0UkxAKxPyyWzQMO0IiIiZ0MBT5qVuFYBDEhshWHA5z9muLscERGRFkkBT5qda/u2BeCzbQp4IiIiZ0MBz0OMGTOGVq1acdNNN7m7lHN2de8YvMwmfjpUQOqRYneXIyIi0uIo4HmIqVOn8u9//9vdZbhEeKCP46LH6sUTERFpOAU8DzF06FCCg4PdXYbLXFc9TPvxlnSsujetiIhIg3hEwFu9ejWjR48mNjYWk8nEp59+2ij7NFZts2fPJjExET8/P5KTk9m4cWOj1NKSXNUrmlB/bw7ll7H6lyPuLkdERKRF8YiAV1JSQlJSErNnz260fdauXUtlZWWt9Tt27CAr69QX5T3T5yxcuJBp06bx9NNPs3XrVpKSkhgxYgTZ2cfvx9q3b1969epVa8nI8NzhSz9vCzdeGAfAfzYccHM1IiIiLYuXuwtwhZEjRzJy5MhG28dmszF58mQ6d+7MggULsFgsAOzevZthw4Yxbdo0Hn300bP6nJkzZzJp0iTuvPNOAN58802++OIL3n77bR5//HEAtm3b1oCWeY5xye14e20aK3ZlkZFfRmyYv7tLEhERaRE8ogevsZnNZhYvXsz333/PHXfcgc1mIzU1lWHDhnH99defMtydSUVFBVu2bCElJcXps1JSUli3bp2ryncye/ZsevTowYABAxrl+K7UqU0Qye3DsRmwcNNBd5cjIiLSYijg1VNsbCwrVqxgzZo1jBs3jmHDhpGSksKcOXPO+pg5OTlYrVaioqKc1kdFRZGZmdmgY6WkpHDzzTezePFi4uLiThkQJ0+ezI4dO9i0adNZ192Ubr8oAYAFmw5QUWVzczUiIiItg0cM0TaVdu3aMX/+fC677DI6dOjAW2+9hclkcndZACxbtszdJTSKET2jiAz2JauwnP/+kMFN/eLcXZKIiEizpx68BsjKyuKee+5h9OjRlJaW8tBDD53T8SIiIrBYLLUmaWRlZREdHX1Ox/YUvl4WJl7cHoB/rErFpkumiIiInJECXj3l5OQwfPhwunfvzqJFi1i+fDkLFy5k+vTpZ31MHx8f+vXrx/Llyx3rbDYby5cvZ9CgQa4o2yPcflE7gn292JNdzIpd2WfeQURE5DznEQGvuLiYbdu2OWabpqWlsW3bNg4csF9e44033mD48OEN2udENpuNkSNHkpCQwMKFC/Hy8qJHjx4sXbqUd955h1deeeWsa5s2bRr/7//9P+bNm8fOnTu57777KCkpccyqFQjx82bcRe0A+MfqVDdXIyIi0vx5xDl4mzdv5vLLL3e8njZtGgDjx49n7ty55OTkkJqa2qB9TmQ2m3nuuecYMmQIPj4+jvVJSUksW7aMyMjIs67t1ltv5ciRIzz11FNkZmbSt29flixZUmvixflu4sXteWfNPjbtO8qGX3NJ7tDa3SWJiIg0Wx4R8IYOHYphnPrcrBkzZjBjxowG7XOyK664os71F1xwwTnVBjBlyhSmTJlS71rOR1EhftzcP47/bDjA377+hYW/vajZTHARERFpbjxiiFbOD1OGdcLHy8zGfXms3pPj7nJERESaLQU8aTFiQv35v2T7dfH+9vXuBvXAioiInE8U8KRFuf/yjgT4WPgxvYCvfj71PYBFRETOZw0OeImJifzpT3+qc7apSGOLCPJ1XBfvhS93Ul5ldXNFIiIizU+DA96DDz7IokWL6NChA1dccQULFiygvLy8MWoTqdO9QzsSEeTLvtxS/v3dfneXIyIi0uycVcDbtm0bGzdupHv37vzud78jJiaGKVOmsHXr1saoUcRJkK8Xj47oCsBry/eQW6z/YIiIiJzorM/Bu/DCC3nttdfIyMjg6aef5l//+hcDBgygb9++vP322zoBXhrVTf3i6NU2hKLyKl7+ere7yxEREWlWzjrgVVZW8sEHH3Dttdfy8MMP079/f/71r39x44038vvf/57bb7/dlXWKODGbTTw9uicA7288yJb9R91ckYiISPPR4Asdb926lXfeeYf3338fs9nMHXfcwSuvvEK3bt0c24wZM4YBAwa4tFCRkw1IDOfmfnF8uCWdJz/5if/97hK8LZoYLiIi0uDfhgMGDGDPnj3MmTOHQ4cO8fLLLzuFO4D27dtz2223uaxIkVP5/ajuhAf6sCuziLfWpLm7HBERkWahwT14v/76KwkJCafdJjAwkHfeeeesixKpr1aBPjw5qjsPf/gDs5b9woie0bSPCHR3WSIiIm7V4B68mnC3efNm5s+fz/z589m8ebPLCxOprxsubMslnSI4VmnjkQ9/wGrTBB8RETm/NTjgpaenM2TIEAYOHMjUqVOZOnUqAwcO5JJLLiE9Pb0xahQ5LZPJxAs39ibI14vN+4/yzloN1YqIyPmtwQHv7rvvprKykp07d5KXl0deXh47d+7EZrNx9913N0aN0pL88hXmLx8h5ujGJv3YuFYB/H5UdwBe+mo3e7OLm/TzRUREmpMGB7xVq1YxZ84cunbt6ljXtWtXXn/9dVavXu3S4qQFSt+MZes7RBbvaPKPHjswniGdIyivsjF1wfe6jZmIiJy3Ghzw4uPjqaysrLXearUSGxvrkqKkBQuNA8C/IrfJP9pkMvHyzUm0CvDm54xC/vb1L01eg4iISHPQ4ID30ksv8bvf/c5pYsXmzZuZOnUqL7/8skuLkxbIEfDy3PLxUSF+/PXGPgD8c/WvfLvniFvqEBERcacGB7wJEyawbds2kpOT8fX1xdfXl+TkZLZu3crEiRMJDw93LHIeCo0HIKCy6XvwalzZM5pxye0AeHDBNjILjrmtFhEREXdo8HXwZs2a1QhliMcIbQuAt7WUymOF4N3aLWU8dU0Pvj+Qz87DhUx5byvv33OR7nIhIiLnjQYHvPHjxzdGHeIpfAIx/MMxleVB4SEIdk/A8/O2MOf2Cxn9+ho27z/KC1/u4o/X9HBLLSIiIk2twQEP7BMqPv30U3bu3AlAz549ufbaa7FYLC4tTlqokLZQloepMB3a9nFbGYkRgbx0cxL3vruFt9ak0SMmhBv7xbmtHhERkabS4DGrvXv30r17d+644w4WLVrEokWL+L//+z969uxJampqY9QoLYxRPdHCVHDQzZXAVb2imXJ5JwCe+OQnth3Md29BIiIiTaDBAe+BBx6gY8eOHDx4kK1bt7J161YOHDhA+/bteeCBBxqjRmlhjJDqXrLCQ+4tpNq0K7qQ0j2Kiiob9/x7M+lHS91dkoiISKM6qwsdv/jii06zZFu3bs0LL7zAqlWrXFqctFDVEy1MBc3j1nVms4lXbk2ia1Qw2UXljH97I0dLKtxdloiISKNpcMDz9fWlqKio1vri4mJ8fHxcUpS0bM2tBw8g2M+bd+4cQEyoH6lHSrhr3ibKKnSnCxER8UwNDnjXXHMN99xzDxs2bMAwDAzDYP369dx7771ce+21jVGjtDSOc/CaRw9ejdgwf+ZNHEiInxdbD+Tzu/e3UmW1ubssERERl2twwHvttdfo2LEjgwYNws/PDz8/Py6++GI6derEq6++2hg1SgtjhNiHaCk6DNYq9xZzki5Rwbw1YQC+XmaW7czmD59uxzAMd5clIiLiUg26TIphGBQWFrJgwQIOHTrkuExK9+7d6dSpU6MUKC1QUBQ2kwWzYYXiTEePXnMxIDGc18ZewH3vbmHBpoOEBnjz+FXdMJlM7i5NRETEJRoc8Dp16sTPP/9M586dFeqkbiYzZd7hBFYcgYL0ZhfwAEb0jObP1/fiyU+2849Vv2KzGfx+VHeFPBER8QgNGqI1m8107tyZ3Fz33WdUWoYyn+o7WDSz8/BOdHtyAn+6ricA/+/bNJ753w4N14qIiEdo8Dl4L7zwAo888gjbt29vjHrEQ5TV3IO2GVzs+HTuGJTIc2N6AzD3u3384dPt2GwKeSIi0rI1+FZld9xxB6WlpSQlJeHj44O/v7/T+3l5eS4rTlqu0hbQg1djXHI7vCwmHvv4R/6z4QBVVoNnx/TCy9Lg//+IiIg0Cw0OeK+88orOU5IzaglDtCe6pX883hYTD3/wAws3HySr6Bivj72AYD9vd5cmIiLSYA0OeBMmTGiEMsTTOAJefvMeoj3RmAvi8Pf24sGF37Ny9xFufnMdb00YQNsw/zPvLCIi0ow0eAzKYrGQnZ1da31ubi4Wi8UlRUnLV+oTaX9ydB+0oIkLV/WKZuE9g4gM9mVXZhHXvbGWbQfz3V2WiIhIgzQ44J1qlmF5ebluVSYOpT4RGJigsgRKjri7nAZJig/j08kX0y06mJzicm79xzo+29Z8brsmIiJyJvUeon3ttdcAMJlM/Otf/yIoKMjxntVqZfXq1XTr1s31FUqLZDN7Q0hbKEyHvDQIauPukhqkbZg/H903mN+9t5Vvdh9h6oJtrP81j6dH90D91CIi0tzVO+C98sorgL0H780333QajvXx8SExMZE333zT9RVKi2WEt8dUmA55v0K7ZHeX02BBvl78vzv6M2vZHmav3Mv7Gw/w/YGjzLqlj7tLExEROa16B7y0tDQALr/8chYtWkSrVq0arSjxEGGJwLdwNM3dlZw1L4uZ6SO6ktwhnIcWbmNXZhE3vLmeMe1MjGxB5xaKiMj5pcHn4H3zzTcKd1IvRnh7+5O8lhvwagzpHMniB4YwqENrSius/Gevhcnv/0B24TF3lyYiIlJLgy+TYrVamTt3LsuXLyc7Oxubzeb0/ooVK1xWnLRsRquagPerewtxkTYhfrx7dzKvL/+F11bsYenObDakreIP1/Tg5n5xuj6kiIg0Gw0OeFOnTmXu3LlcffXV9OrVS7/U5JSMsET7kxY8RHsyi9nE5KEd8Dmyi8U54WzPKOTRj37kfz9k8NyY3sSHB7i7RBERkYYHvAULFvDBBx8watSoxqhHPEmrRPtjaS4cKwC/ULeW40ptA+HDGwYyb0M6M5f+wrd7chg+cxUTL27P5Ms76g4YIiLiVg0+B8/Hx4dOnTo1Ri3iaXyDIbD6gscecB7eybwsZu69rCNLpg5hcMfWVFTZeHNVKkNfWsl/Nuynymo780FEREQaQYMD3sMPP8yrr756ygseizgJ72B/9KBh2pN1iAziP3cn8687+tMhIpDckgqe/GQ7I1/9ls9/zMBq08+KiIg0rQYP0a5Zs4ZvvvmGL7/8kp49e+Lt7TwUtWjRIpcVJx6gVXs4uMFjJlqcislkIqVHFJd1jeTd9fuZtWwPe7KLmfLe93Rqs4ffDevENX1isZh1zqqIiDS+Bge8sLAwxowZ0xi1iCfyoEul1Ie3xcydF7fnhgvieHttGm+vTWNvdjFTF2xj1rI93HNpB67v2xZ/H90PQ0REGk+DA94777zTGHWIp6oZoj1PAl6N0ABvHrqiC3cNac+/v9vHv9akkZZTwhOLfuKFL3dx64B4fnNRgmbdiohIo6j3OXjZ2dmnfb+qqoqNGzeec0HiYWquhefB5+CdToifN1OGdWbtY8P4w9XdiQ/3p6Cskn+u/pVLX/qGu+dt4qufM6mo0oQMERFxnXoHvJiYGKeQ17t3bw4ePOh4nZuby6BBg1xbnbR8NT14hYegssy9tbhRoK8Xdw/pwMrpl/PW+P4M6RyBYcCyndn8dv4Wkp9bxtOfbeeHg/mawCQiIues3kO0J//S2bdvH5WVlafdRoSAcPv1744VQG4qRPdyd0VuZTGbGN49iuHdo9ibXcyHmw/yyfeHyC4qZ966/cxbt5/E1gGM6BnNiF7R9I0Lw6yJGSIi0kANPgfvdHRXC6nFZIKIrpC+EXJ2n/cB70Sd2gTxxKjuPHpVN9buzWHR1nSW/JzJvtxS/rH6V/6x+leiQny5skc0V/WKZmD7cLwtDb6ykYiInIdcGvBE6tSmmz3gHdnt7kqaJYvZxKVdIrm0SyQl5VWs+uUIS7ZnsmJXNlmF5cxfv5/56/cTFuDNpZ0jGdI5giGdI4kO9XN36SIi0kzVO+CZTCaKiorw8/PDMAxMJhPFxcUUFhYCOB5FaonsZn88ssu9dbQAgb5ejOodw6jeMZRXWfkuNZevtmfy9Y4s8koq+O8PGfz3hwwAOrcJYkjnSIZ0iSC5fTgBPvr/moiI2DXoHLwuXbo4vb7gggucXmuIVuoU2dX+qB68BvH1snB51zZc3rUNz44x2HrgKN/+coTVe3L4MT2fPdnF7Mku5u21aXhbTPSJC2NAYjgDElvRPyGc0ADdD1dE5HxV74D3zTffNGYd4slqevBy94K1EiwKHg1lMZuqw1s4067sSn5pBd+l5vLtniOs/iWHQ/llbNl/lC37j/LmKvs+XaOCGdDeHvb6xoeR0DpA/wkTETlP1DvgXXbZZY1Zh5yjMWPGsHLlSoYPH85HH33k7nKchbQFnyCoKLbPpG3Tzd0VtXhhAT6OoVzDMDiQV8rGtDw27ctj876j/JpTwu6sInZnFfHu+gMAhPp70yculL7xYSTFhdEnPpQ2wTqPT0TEE+mkHQ8xdepUJk6cyLx589xdSm0mk32Y9tAW+3l4CnguZTKZSGgdSELrQG7uHw/AkaJytuzPY2PaUbYcOMrOjEIKyir5dk8O3+7JcewbG+pHUnwYfeLC6B4TTI/YEIU+EREPoIDnIYYOHcrKlSvdXcapRXarDng6D68pRAb7clWvGK7qFQNARZWN3ZlFbEvP58eD+fxQfQ5fRsExMgoy+XJ7pmPfiCAfukYF41tqpnJbBr3jw+kQGahLtIiItCBu/xd79erVjB49mtjYWEwmE59++mm99ps9ezaJiYn4+fmRnJxc6zZpRUVFPPjggyQkJODv78/gwYPZtGmT4/0ZM2ZgMpmclm7dXN+zVN/2nak9LZ5jooVm0rqDj5eZ3nGh/OaiBF66OYmvH7qMn2aMYME9F/HEyG5c0yeGjpGBmE2QU1zB2tRcVhw2M/3j7YyYtZqeT33F1a99y/QPf+CtNWms3ZvDkaJyXdxcRKSZcnsPXklJCUlJSUycOJEbbrihXvssXLiQadOm8eabb5KcnMysWbMYMWIEu3fvpk2bNgDcfffdbN++nfnz5xMbG8u7775LSkoKO3bsoG3btgD07NmTZcuWOY7r5XX6P461a9cycOBAvL2dJwns2LGD1q1bExUVdVbtq097+vbtS1VVVa19v/76a2JjY09bd7MQ2d3+qB68ZiPI14uLOrTmog6tHevKKqzszipie/pRvlq/nVLfcHZnFVNcXsXPGYX8nOF8OaRWAd50iQq2L9HBdGkTRJeoYFoF+jR1c0RE5ATnHPAKCwtZsWIFXbt2pXv37g3ef+TIkYwcObJB+8ycOZNJkyZx5513AvDmm2/yxRdf8Pbbb/P4449TVlbGxx9/zGeffcall14K2Hvs/ve//zFnzhz+8pe/APZAFx0dXa/PtNlsTJ48mc6dO7NgwQIsFgsAu3fvZtiwYUybNo1HH330rNp3pvYAbNu2rV51Nls1PXi5e8BaBRa3/99C6uDvY6FvfBg9owMJzv6RUaMGYrF4kX60jB2HC9hxuIidhwvZk1XE/rxSjpZWsiEtjw1peU7HiQz2pUtU0PHwFxVMl6gggv00g1pEpCk0+LfsLbfcwqWXXsqUKVMoKyujf//+7Nu3D8MwWLBgATfeeGNj1OlQUVHBli1beOKJJxzrzGYzKSkprFu3DoCqqiqsVit+fs4ni/v7+7NmzRrH6z179hAbG4ufnx+DBg3i+eefp127dnV+rtlsZvHixVx66aXccccdzJ8/n7S0NIYNG8b1119fZ7hzVXtcafbs2cyePRur1eryY59WaDx4B0BlKRxNg4jOTfv5ctbMZhPtWgfQrnWA45w+sPf27c0u5pesIn7JLuKXzCJ+ySrmUH4ZR4rKOVJUztq9uU7Hig31o3NUMJ3aBDmWjpFBhKvHT0TEpRoc8FavXs2TTz4JwCeffIJhGOTn5zNv3jz+8pe/NHrAy8nJwWq11hoOjYqKYtcu+/ldwcHBDBo0iD//+c90796dqKgo3n//fdatW0enTp0ASE5OZu7cuXTt2pXDhw/zzDPPMGTIELZv305wcHCdnx0bG8uKFSsYMmQI48aNY926daSkpDBnzpxGbU99pKSk8MMPP1BSUkJcXBwffvghgwYNqrXd5MmTmTx5MoWFhYSGhp513Q1mNkNEFzi8DbJ3KuB5AH8fC73jQukd5/z3qLi8ij1ZRfbgl1UdALOKyCosr57UcYxVvxxx2ic80IdOkUF0bBNIx8jjwa9tmD9ms67dJyLSUA0OeAUFBYSHhwOwZMkSbrzxRgICArj66qt55JFHXF7g2Zo/fz4TJ06kbdu2WCwWLrzwQsaOHcuWLVsAnIZN+/TpQ3JyMgkJCXzwwQfcddddpzxuu3btmD9/PpdddhkdOnTgrbfeahYXjz3xXMJmK6qnPeBlbYce17q7GmkkQb5eXNCuFRe0a+W0vqC00t7Tl1XE3uxiUo+UkJpt7/HLK6lgY0keG/c5D/X6e1voEHk89NUEv8SIAHy9LE3ZLBGRFqXBAS8+Pp5169YRHh7OkiVLWLBgAQBHjx6tNSTaGCIiIrBYLGRlZTmtz8rKcjqfrmPHjqxatYqSkhIKCwuJiYnh1ltvpUOHDnUeNywsjC5durB3797Tfn5WVhb33HMPo0ePZtOmTTz00EO8/vrrjd4ejxCTBNv+Axnb3F2JuEFogLfjbhwnKimvIi2nhL3ZxdXBz/64L7eEskprnZM7LGYT7cID6BgZSMfq0NchIpAOGu4VEQHOIuA9+OCD3H777QQFBZGQkMDQoUMB+9Bt7969XV1fLT4+PvTr14/ly5dz/fXXA/YJEMuXL2fKlCm1tg8MDCQwMJCjR4/y1Vdf8eKLL9Z53OLiYlJTU/nNb35zys/Oyclh+PDhdO/enQ8//JBffvmFoUOH4uvry8svv9wk7WnRYvraHw//4NYypHkJ9PWiV9tQerV1HuqttNo4mFfq6O3bm13M3iPFpGbbZ/Wm5ZSQllPCsp3ZTvuFBXjTPiKQDhFBdIgMdAS/hNYB+Hmr109Ezg8NDnj3338/AwcO5ODBg1xxxRWYzfZL6XXo0MExO7UhiouLnXrN0tLS2LZtG+Hh4bRr14433niDTz75hOXLlzu2mTZtGuPHj6d///4MHDiQWbNmUVJS4piFCvDVV19hGAZdu3Zl7969PPLII3Tr1s2xzfTp0xk9ejQJCQlkZGTw9NNPY7FYGDt2bJ112mw2Ro4cSUJCAgsXLsTLy4sePXqwdOlShg0bRtu2bXnooYca3L76tscjRPcCkxmKM6EoE4I9rIdSXMrbYqZDZBAdIoOc1huGQXZRuVNv369HSvj1iP3CzfmllXx/IJ/vD+Q77WcyQdswfzpEBpEY7k9plonQ1Fy6RIcSHeKnc/1ExKOc1bUq+vfvT//+/QGwWq389NNPDB48mFatWp1hz9o2b97M5Zdf7ng9bdo0AMaPH8/cuXPJyckhNTXVaZ9bb72VI0eO8NRTT5GZmUnfvn1ZsmSJ00SFgoICnnjiCdLT0wkPD+fGG2/k2WefdVzDLj09nbFjx5Kbm0tkZCSXXHIJ69evJzIyss46zWYzzz33HEOGDMHH5/gQUFJSEsuWLTvlfmdqX33b4xF8Au0TLY7ssvfiKeDJWTCZTESF+BEV4sfFnSKc3iursJKWU8KvOcWkHSnh1xx78Pv1SAlF5VWkHy0j/WgZqwGw8FGa/ZxcP28z7SNqhnntS/vqHsAQXdpFRFqgsxqi7d27N3fddRdWq5XLLruM7777joCAAD7//HPHkG19DR069LRXw58xYwYzZsyotX7KlCmnHcK85ZZbuOWWW075fs25gw1xxRVX1Ln+ggsuOOU+Z2pfjTO1x2PEJB0PeF1GuLsa8TD+PhZ6xIbQIzbEab1hGOQUV/DrkWLSckrYk1XIhh1plFqCOJBXxrFKGzsPF7LzcGGtY0YE+dYZ/NqFB+j2bSLSbDU44H300Uf83//9HwD/+9//SEtLY9euXcyfP58nn3yStWvXurxI8SAxfeHHhZpoIU3KZDIRGexLZLAvyR1aU1lZyWJbKqNGXQJmC+lHyxw9fY5ev5wSjhSVk1NsX06e4Wsxm4hr5U9i60DaRwSS2DqAxAj787Zh/ngp/ImIGzU44OXk5Dhmdy5evJibb76ZLl26MHHiRF599VWXFygeJibJ/qiJFtJMeFvMtK8OZsNPuhlP0bFK+5DvScO9aTn2Gb77c0vZn1ta67p+3hYT8a3sgc8eAI8/jw3zx6Lz/USkkTU44EVFRbFjxw5iYmJYsmSJ4yK/paWljtt3iZxSdPVM68J0KMmBwIjTby/iRsF+3vSJC6NPXJjTesMwyCw8RlpOCftyStmXW1L9vIT9eaVUVNnsgTCnpNYxfbzMtAsPcAp+7VsHkhgRqMkeIuIyDQ54d955J7fccgsxMTGYTCZSUlIA2LBhA926dXN5geJh/EKgdSfI3Wu/6HGnFHdXJNJgJpOJmFB/YkL9GdzR+T2bzeBw4TH25RwPfTUB8EB1+Ku55t/J/LzNJIQHknhS8GsfEUibYN9mcVF1EWkZGhzwZsyYQa9evTh48CA333wzvr6+AFgsFh5//HGXFygeKCapOuD9oIAnHsdsNtE2zJ+2Yf61ZvlabQYZ+WX24Jd7YgAs5WBeKccqbezOKmJ3VlGt4wb4WEio6fU7Ifgltg4kIshH4U9EnJzVZVJuuummWuvGjx9/zsXIeSKmL2z/GNK3uLsSkSZlMZuIDw8gPjyAS3G+tFKl1caho2Wk5VaHvpwS0nJL2ZdTQvrRUkorrKec6RvoY6Fd60ASwgNIiAggITyQtqE+5JXbQ6Uu9CJy/jmrgLdq1Spefvlldu7cCUCPHj145JFHGDJkiEuLEw/VbpD98cA6sNnArNmGIt4Ws30iRkQgdHV+r6LKxsGjpceHfXPt5/6l5ZSQUVBGySnDnxfP/bCM+PAAe/hrHUhC6wASWgfQLjyQ+HB/3dNXxEM1OOC9++673Hnnndxwww088MADAKxdu5bhw4czd+5cxo0b5/IixcPE9gXvACjLs18TL6qHuysSadZ8vMx0jLTfc/dk5VVW0o+WsT+3xDGr1/7cvlRaqb7TRwngPNvXZILYUH/7pI8Ie+hLbB1Au9b2MBjke1Z9ACLSDDT4p/fZZ5/lxRdfdLot1wMPPMDMmTP585//rIAnZ2bxhviB8OtK2L9WAU/kHPh6WeoMf5WVlXz+xWIuuPhyDhVU2INfXgn7c0rZn2cPgaUVVg7ll3Eov4x1v+bWOnZEkA/tTur5S6geCg4P1Hl/Is1ZgwPer7/+yujRo2utv/baa/n973/vkqLkPJBw8fGAN3CSu6sR8Ujm6vvvJkaGcHEn5/dq7u5xIM8+3Ls/r5QDufYJHwfySskrqSCn2L5sPem+vgBBvl5Ow72JrQNoV31+YUyony70LOJmDQ548fHxLF++nE6dnP+1WLZsGfHx8S4rTDxcwmD74/7vwDDsY0Ui0mROvLtHv4TwWu8XHqvkQPWQ777cEg7UPOaVcrjgGMXlVfycUcjPGbUnfXiZTbRt5U98K3vga3fSEhqgaR8ija3BAe/hhx/mgQceYNu2bQwebP8lvXbtWubOnas7WUj9te0PFh8ozoK8X6F1xzPvIyJNJsTPm15tQ+nVNrTWe8cqrRzMOyH8VT8/eLSU9LwyKqw2x/mAdR/bi3Y1PX4nhcDYMH98vNT7J3KuGhzw7rvvPqKjo/nb3/7GBx98AED37t1ZuHAh1113ncsLFA/l7WcPeQe+g31rFPBEWhA/bwudo4LpHBVc6z2bzSCr6BgHqod6D+ZVPx4t40BeKUeKyik8VsX2Q4VsP1S7989sgpjqiR/twu0TPuJaHX+tc/9E6qdBAa+qqornnnuOiRMnsmbNmsaqSc4XCYPtAW//d9BP11EU8QRm8/G7fCR3aF3r/dKKKtKPljkC4IG8UtKPHn9+rNJ22okfgT4WR49fXJgfhZkmAn85Qvs2IbQN88fPW5d9EYEGBjwvLy9efPFF7rjjjsaqR84niRfDty/be/B0Hp7IeSHAx4suUcF0qaP3zzAMjhSXO3r9DuSWcfDo8Z7AzMJjlFRY2ZVZxK7Mmrt9WPgo7XvHMaJD/ByTPew9gP6O15FBut2bnD8aPEQ7fPhwVq1aRWJiYiOUI+eV+IvA4guF6ZC9U5dLETnPmUwm2gT70SbYr86JH8cq7Zd1qQl8+3OK2bQzjQqfUA7mlVJSYSWz8BiZhcfYuC+v1v6+XmbiWvkTH24f9o1rZX+Mr37U8K94kgYHvJEjR/L444/z008/0a9fPwIDA53ev/baa11WnHg4nwBoPwT2LoM9Xyngichp+Xk7X/OvsrKSxbZURo0ahJeXF0dLKx1DvQdPOP/vQF4pGflllFfZSD1SQuqRkjqPH+BjcQp+x8Of/XVYgLcCoLQYDQ54999/PwAzZ86s9Z7JZMJqtZ57VXL+6DyiOuAthUseOvP2IiJ1MJlMhAf6EB7oQ9/4sFrvV1ptHM4/RvrRUtKP2od+04+WkX60lIN5ZWQVHaO0wsovWcX8klVc52cE+Xo5gt/xEBjg6BUM9dflX6T5aHDAs9lsjVGHnK+6XAlfPgIH1kPZUfBv5e6KRMQDeVvM9kuztA6o8/3yKiuH84/VCn41gTC7qJzi8qqTzv9zFuznRVyrAOJP7gWsHhIO9lMAlKajGw2Ke7VKhMhu9nvSpq6AXje6uyIROQ/5ellIjAgkMSKwzvdrzv+rK/ylHy0lp7iComNV7DxcyM7DtS//AhDq7018uD9xYSeHP/vrQN37V1yo3n+bVqxYwZQpU1i/fj0hISFO7xUUFDB48GDmzJnDpZde6vIixcN1vtIe8H75WgFPRJqlk8//O1lZhZVD+ScHv+NDwXklFRSUVVJwqLLO6/8BhAf6OA0BRwf7kJFnov3hIhIiggnx99I5gFJv9Q54s2bNYtKkSbXCHUBoaCi//e1veeWVVxTwpOG6jIDvXoO9S8FmBbOuYyUiLYu/j4VObYLp1Kb25V8ASsqrHL196UfLOJhXHQLz7Y/5pZXklVSQV1LBj+kFJ+xp4V+71wH2cwDbhvkTG+ZH21b+tA0LIDbMj7jq55HBvljMCoBiV++A98MPP/DXv/71lO9feeWVvPzyyy4pSs4z8cngGwqluXBww/H71IqIeIhAXy+6RgfTNbruAFh4rJJDNb1+NeHvaAk792dRgg95JZUUl1exO6uI3Vl1nwPobTERHepXHQL9iQvzp20r+/OadboQ9Pmj3gEvKysLb+9TnyDq5eXFkSNHXFKUnGcs3tDtavjhPdi+SAFPRM47IX7ehMR40z3m+ChZZWUlixcvZtSoy6kyzBzKLyOj+i4fGfll9kBY/Tyz4BiVVoODeWUczCs75edEBPk4wl7bkwJgXCt/Qv11KRhPUe+A17ZtW7Zv306nTp3qfP/HH38kJibGZYXJeabXjfaAt+NTuOoFsOhkYxGRGvYh4CA6tan7HECrzSCr8JgjAKYfrR0GSyqs5BRXkFNcwQ9Ow8DHBfpY7IHvpOBX8zwqxE/DwC1EvX+Ljho1ij/+8Y9cddVV+Pn5Ob1XVlbG008/zTXXXOPyAuU80eEy8A+HkiOwbzV0HObuikREWgyL2URsdc9c/zreNwyDgrJK+31+j5Y59QbaXx8jp7ickgore7KL2ZNd97UALWYT0SE15wDW7gVsG+aPv4+GgZuDege8P/zhDyxatIguXbowZcoUunbtCsCuXbuYPXs2VquVJ598stEKFQ9n8Yae18Pmt2H7xwp4IiIuZDKZCAvwISzAh56xoXVuc6zSyuGCY9WBr5RD+cefZ+Qf43BBGZVWwx4K8089DBwe6ENsmB8xof7EhvoRE+ZPTKgfsdWPUSF+eFvMjdVUqVbvgBcVFcV3333HfffdxxNPPIFhGID9L82IESOYPXs2UVFRjVaonAd63WgPeDv/B1fPBC9fd1ckInLe8PO20D4ikPanuBag1WZwpKjcEfAyTu4NPFpGUXmVYzbwqS4HYzJBm2BfYkL9iQ7x5Viemazv9hMXHugIghFBmhF8rhp0olNCQgKLFy/m6NGj7N27F8Mw6Ny5M61a6e4D4gLtBkFwDBQdhr3Lodsod1ckIiLVLGb7LN3oUD/6JdT9e7+gzD4b+HBBGRkFxzicX8bhAvu5gYcLjpFZcIwKq42swnKyCsur9zKz8vBup+N4mU1Ehfg5egJjwvyIDXXuCQwP9NGEkNM4qzPZW7VqxYABA1xdi5zvzBZ7L966N+D7+Qp4IiItTKi/N6H+3vSIrX3NXACbzSC3pMIeAPOPkZ5XzHff78S/dSyZReUczi8jq6icKtuJQ8FH6zyWr5eZmNCTAuAJjzGh/oT4nb8Xh9ZURWleLrzDHvB+WQKFGRAS6+6KRETERcxmE5HBvkQG+9Inzn4pmMijPzNqVB/HpdiqrDaOFJc7zvs7nH+MjOrHmp7BI0XllFfZ2Jdbyr7c0lN+XqCP5fg5gLUCoB/Rof4Eeegt4jyzVdJyRXaFhIth/1r4/l247FF3VyQiIk3Iy2K298qF+gN1DwVXVNkcl4U5XHA8AGbkVw8NF9jvDlJSYWVvdjF7TzErGOx3CIkO9SM6xM/psWZCSEsdDlbAk+an3wR7wNsyD4Y8rFuXiYiIEx8vM/HhAcSHB5xym7IK6/Hgd1IP4OH8MjILj1F0rIri8qozhkAfi5moUN/q8GefHGJ/9HOcl9gm2LdZzQ5WwJPmp/u14P8oFKbbJ1t0udLdFYmISAvj72OhY2QQHSPrvjg02O8RnFlon/yRWXDM8fxwwTGyCu2PuSXlVFhtJ9wlpO5zAk0miAzydeoFHNatDRd3cM9EVAU8aX68/SBpHKyfDZv+pYAnIiKNItDX64whsKLKRnbR8cBXVxjMLrLfKi67qJzsonJ+xH6nkIggXwU8EScD7oL1f4c9X0H2LmjTzd0ViYjIecjHy0xcqwDiWp16OLhmdrAjBBYeI7OgjOT24U1YqTMFPGmeWneE7tfYL3r83etw/Wx3VyQiIlKnE2cH92rrfKeQyspK99Tklk8VqY+LH7Q//rjQfskUERERqRcFPGm+4vpDu8Fgq4T1c9xdjYiISIuhgCfN28VT7Y+b34HSPPfWIiIi0kIo4Enz1vlKiOoNFUXw3WvurkZERKRFUMCT5s1shmFP2p9v+AcUZ7u3HhERkRZAAU+avy5XQdt+UFkKa15xdzUiIiLNngKeNH8mEwz7g/35pregIN299YiIiDRzCnjSMnS4HBIuAWs5LP+zu6sRERFp1hTwpGUwmeDK6mD34wI4tNW99YiIiDRjCnjScrS9EPrcZn/+1ZNgGO6tR0REpJlSwJOWZfgfwcsfDnwHOz5zdzUiIiLNkgKetCyhcXDxA/bnX/0eyovdW4+IiEgzpIAnLc8lD0FYAhQegpXPu7saERGRZkcBT1oeb38Y9bL9+fo5kLndvfWIiIg0Mwp40jJ1uRK6jwbDCp8/CDaruysSERFpNhTwpOW66gXwCYb0TbDhTXdXIyIi0mwo4EnLFRp3/Np4y/8MuanurUdERKSZUMCTlq3fBGh/KVSVwX9/BzabuysSERFxOwU8adlMJrj2dfAOhP1rYf1sd1ckIiLidgp4HmDMmDG0atWKm266yd2luEerRLjqOfvz5X+CzJ/cWo6IiIi7KeB5gKlTp/Lvf//b3WW414XjoesosFbAx5OgsszdFYmIiLiNAp4HGDp0KMHBwe4uw71qhmoD28CRnfa7XIiIiJynmn3AW716NaNHjyY2NhaTycSnn35ar/1mz55NYmIifn5+JCcns3HjRqf3i4qKePDBB0lISMDf35/BgwezadMmt9R+plqlngIjYMybgAk2vw0/feTuikRERNyi2Qe8kpISkpKSmD27/ifPL1y4kGnTpvH000+zdetWkpKSGDFiBNnZ2Y5t7r77bpYuXcr8+fP56aefuPLKK0lJSeHQoUN1HnPt2rVUVlbWWr9jxw6ysrLOuvb61Nq3b1969epVa8nIyKjvH8n5o9NwGPKw/fn/pkLOXvfWIyIi4gbNPuCNHDmSv/zlL4wZM6be+8ycOZNJkyZx55130qNHD958800CAgJ4++23ASgrK+Pjjz/mxRdf5NJLL6VTp07MmDGDTp06MWfOnFrHs9lsTJ48mXHjxmG1Hr9jwu7duxk2bBjz5s0769rPVCvAtm3b2L59e60lNja23n8m55WhT0DCxVBRDAtvh/Iid1ckIiLSpJp9wGuoiooKtmzZQkpKimOd2WwmJSWFdevWAVBVVYXVasXPz89pX39/f9asWVPrmGazmcWLF/P9999zxx13YLPZSE1NZdiwYVx//fU8+uijjVarq82ePZsePXowYMCARjl+s2DxgpvehqBoOLILFv1W18cTEZHziscFvJycHKxWK1FRUU7ro6KiyMzMBCA4OJhBgwbx5z//mYyMDKxWK++++y7r1q3j8OHDdR43NjaWFStWsGbNGsaNG8ewYcNISUmps8fPlbXWR0pKCjfffDOLFy8mLi7utOFw8uTJ7Nixw+XnGzY7wdFw23/A4gO7v4CVz7m7IhERkSbjcQGvvubPn49hGLRt2xZfX19ee+01xo4di9l86j+Sdu3aMX/+fBYuXIiXlxdvvfUWJpOpCauu27Jlyzhy5AilpaWkp6czaNAgd5fUPMT1h9Gv2p+vfgm21D2ULiIi4mk8LuBFRERgsVhqTXzIysoiOjra8bpjx46sWrWK4uJiDh48yMaNG6msrKRDhw6nPHZWVhb33HMPo0ePprS0lIceeqhJapVz0HccDJluf/75Q7B7iXvrERERaQIeF/B8fHzo168fy5cvd6yz2WwsX768zp6twMBAYmJiOHr0KF999RXXXXddncfNyclh+PDhdO/enUWLFrF8+XIWLlzI9OnTm6xWOUvD/gBJ48CwwocTIH2zuysSERFpVF7uLuBMiouL2bv3+KUu0tLS2LZtG+Hh4bRr14433niDTz75xCkkTZs2jfHjx9O/f38GDhzIrFmzKCkp4c4773Rs89VXX2EYBl27dmXv3r088sgjdOvWzWmbGjabjZEjR5KQkOAYnu3RowdLly5l2LBhtG3bts7evDPVXt9a5RyZTHDta1CcBanL4b1b4K6l0LqjuysTERFpFM0+4G3evJnLL7/c8XratGkAjB8/nrlz55KTk0NqaqrTPrfeeitHjhzhqaeeIjMzk759+7JkyRKnyQwFBQU88cQTpKenEx4ezo033sizzz6Lt7d3rRrMZjPPPfccQ4YMwcfHx7E+KSmJZcuWERkZeVa117dWcQGLN9zyb5h7NRzeBvPHwF1f2ydjiIiIeJhmH/CGDh2KYRinfH/GjBnMmDGj1vopU6YwZcqUU+53yy23cMstt9S7jiuuuKLO9RdccMEp9zlT7TXOVKu4iG8Q3P4hvHUFHN0Hc6+BCZ8r5ImIiMfxuHPwRE4rqA385lMIjYfcPfYevcK6L40jIiLSUingyfknvL295y40HnL3Voc83fZNREQ8hwKenJ9aJVaHvHaQl2ofrlXIExERD6GAJ+evmpAXVh3y3r4Kcva4uyoREZFzpoAn57dWCTDhC2jVHvL32ydgHFjv7qpERETOiQKeSFg7+3Xx2vaDsqMw71rY8Zm7qxIRETlrCngiAEGRMP5z6DoKrOXwwXhYN9vdVYmIiJwVBTyRGj4BcOu7MOBuwICvfg+fTYbKMndXJiIi0iAKeCInMltg1MtwxZ/AZIbv34V/pUBu6pn3FRERaSYU8EROZjLBxVPtF0QOjISs7fCPy+DnT91dmYiISL0o4ImcSofL4LffQrvBUFEEH46HL6ZDZam7KxMRETktBTyR0wmJgfH/g4sftL/e9P/w+tdQwot/cWtZIiIip6OAJ3ImFi+44hn4v48hOBZT3q9csudZzMue0gQMERFplhTwROqrUwrcvw5bn3GYMLBs+Du8OQTSvnV3ZSIiIk4U8EQawj8M6+jXWN9hGkZQFOTugXnX2K+bl3/A3dWJiIgACngiZyUrtC9V96y1XzPPZIYdn8IbA2DlCxq2FRERt1PAEzlb/mFw9d/sM20TLoGqY7DyeXi9H2x6C6oq3F2hiIicpxTwRM5VdC+Y8DncPBdC46HwEHwxDV6/ELbMBWuluysUEZHzjAKeiCuYTNBzDEzZDCNfgqBoKDgI/5tqD3ob/gnlxe6uUkREzhMKeCKu5O0HyffA1G1w1QsQ2MY++eLLR+CVHvD1H6Eg3d1VioiIh1PAE2kM3v5w0X0w9Qf7vW3DO8CxAvjuNZjVBz6cAHuXg83q7kpFRMQDKeCJNCafABg4CaZsgbELIHEIGFb4+RN49wZ4NQm+eQ6O7nN3pSIi4kG83F2AyHnBbIauI+1L5k+w9d/w4wf28/RW/dW+xF9kP4+vx7UQEuvuikVEpAVTD55IU4vuDaNegod3w01vQ8dhgAkOroclj8HMHvD2SPvEjKJMd1crIiItkHrwRNzF2w963WhfCjNgx3/tQ7cH18OB7+zLl49CwmDochV0vhIiu9pn7IqIiJyGAp5IcxASCxfda18KDsGOz+xhL30j7F9rX5b+EULbQecr7GGv/RDwCXR35SIi0gwp4Ik0N6FtYdD99iX/IOz6AvYuhbRvoeAAbH7Lvlh87b177YdA4qUQewFY9CMtIiIKeCLNW1j88Z69ilLY9y3sWQp7vrJfX+/Xb+wLgE8QtBsEiZfYQ190kgKfiMh5Sv/6i7QUPgHQZYR9MV6CnF/g15WQtto+hFt21N7Tt3dp9fZB0PZCiBsI8QMhbgAEhLu1CSIi0jQU8ERaIpPJPuEisisk/xZsNsjaDvvW2Hv59q2F8gJ7+EtbfXy/1p2qA98A+2Ob7mC2uK8dIiLSKBTwRDyB2QwxfezLoPvtd8g4sgsOboT0TfbH3D2Qu9e+/PCefT/vQPtlW2L7QkxfiEmCiC7ubImIiLiAAp6IJzJbIKqnfel/p31daR6kb4aDG+yzcw9thYpi+2VZDq4/vq+XP5aonvQuD8P0Qz7EXQiR3cDi7ZamiIhIwyngiZwvAsKhy5X2Bey9fDl74PA2OPwDZGyDzB+hohjzoc10APh8mX1bi689LEb3gqiapQf4t3JPW0RE5LQU8ETOV2YLtOlmX5Jus6+z2SAvlaqDm9m37jM6+BdizvwJygshY6t9OVFo/PGewprg17qjzusTEXEzBTwROc5shojOGKGJ/HwggIRRozBbLHA0zd7Ll70DMrdD1s/2a/IVHLQvvyw5fgwvP/vkjaieENXb3tMX2R2CIt3XLhGR84wCnoicntls75Vr3RG44fj6snx74Mv6GTJ/sj9m74DKUsj43r6cKKC1Pei16WY/py+ymz0IBkY0ZWtERM4LCngicnb8w+x30kgYfHydzWbv7cuq7uXL3G4PfUf3QWku7F9jX04UEGEPepHVw8WR3e2vdc0+EZGzpoAnIq5zYm9fj+uOr68otV+Y+cguyN55/DF/P5TmVF+771vnYwW2Od7bF9Hl+BIcbb8OoIiInJICnog0Pp8A+7X2Yvs6r68ogSO7Twp+u+zn95VkQ1q284WaAXyCIaKz/SLPEZ2PB7/wDrqUi4hINQU8EXEfn0D77dTaXui8vrwYcnbbw17ObvvlXHJ+gbw0qCiqe0av2QtatbeHvcgTevwiOoNfaNO1SUSkGVDAE5HmxzcI2vazLyeqKreHvJxfnINfzh77RZtz99iX3V847xcU7dTbZ2rVEb+KXDBsTdcmEZEmpIAnIi2Hl+/xa/edyDCgMON42DsxABYdhuJM+1J9np8XMAIwfnkSwqvPGYzobL9Xb83iH9bUrRMRcRkFPBFp+UwmCG1rXzpe7vzescITQp99MY7sxsj9FXNlKWT9ZF9OFhBhD3oRJ4S+1p0hvL09aIqINGMKeCLi2fxCIK6ffalWVVnJl1/8l5GDeuJdsN8eAHP3Hl+KDttn95bmON+nF8Bktt/B4+Qev9adIKStfSaxiIibKeCJyHnJMHnZh2ejukGXEc5vlhdDXmp18EutPrdvL+TstU/yyN9vX/Yuc97Py//4ZWJqevxad7K/1nX9RKQJKeCJiJzMNwhikuzLiQwDirNP6O2rCYB77ZM/qsqqL/K8vfYxA1qf1OPX0R4wwzvYLyMjIuJCCngiIvVlMkFwlH1JvNj5PWuVvVfvxB6/ml6/ogz7nTxKc+HghtrHDY6pnuzRwR74aoKfwp+InCUFPBERV7B4nXDP3iud36sZ8s3daw+ANef85f0Kx/Lt5/wVHa59Gzc4Hv7C2zv3+in8ichpKOCJiDS2Uw35ApTm2YNebqr9MS/1+OtzCX8m3dVD5HymgCci4k4B4fYlrn/t984h/HkFRXMxYVg+/woiOzsP/arnT8TjKeCJiDRX5xD+TMWZRJAJP+yqvW+tnj+FPxFPo4AnItISnSH8VWX/wg/fLKJvuxAs+fsaNuzbqr09/J342CrR/nkmUyM3TERcQQFPRMTTBIRjtO1HengWfS4dhcX7hPPx6jvse+C72sf1DbEHPafwl2h/HhoHZksTNVBEzkQBT0TkfFKfYd+8NDiadvzx6D576CsvhMwf7cvJzN4Q1q7u8NcqUUO/Ik1MAU9EROxOF/4qSu3X+Tsx9Dme7wdbZXVvYGrdxw6Kdh7uPTEIBrTW0K+IiyngiYjImfkEQJvu9uVkNisUHqoOfPuce//y9kF5ARRn2pcD6+o4dnB1b19C9WMihFU/D2sH3n6N2TIRj6SAJyIi58ZssQexsHbAZc7vGQaUHa0d+mpeF2XY7++b9ZN9qUtQtHMArAl/wW3BsDVq00RaKgU8DzFmzBhWrlzJ8OHD+eijj9xdjoiIncl0fOi3bb/a71ceOz70m7/fPtybv7+6J3AfVBQf7/07uN5pV29gtMmC6UAdPX81YdC/VWO3UKRZUsDzEFOnTmXixInMmzfP3aWIiNSftx9EdrUvJzMM+8SP/H3Vge/E8Lcfo+AgZltV9cSQX+s+vm8otGp3Uvirfq7hX/FgCngeYujQoaxcudLdZYiIuI7JBIGt7UsdvX9V5WV889/3GNa3PV5Fh5zCH0f3QUm2/fy/zJ/sS12CY2r3/NUMNwfH2u8xLNICmd1dwOrVqxk9ejSxsbGYTCY+/fTTeu03e/ZsEhMT8fPzIzk5mY0bNzq9b7Va+eMf/0j79u3x9/enY8eO/PnPf8YwDABmzJiByWRyWrp16+bq5tW7fWdqj4iInMTsRZlPBEbiELjwNzDsD3Djv+DupfDIHvj9Ybh/A4xdCCNfhIvuh65XQ5ue4BNkP0bRYfvEjx8XwKoX4NP7YO7VMKs3/KWN/XHuNfDp/bDyBdj2HqR9aw+R1ir3tl/kNNz+X5OSkhKSkpKYOHEiN9xwQ732WbhwIdOmTePNN98kOTmZWbNmMWLECHbv3k2bNm0A+Otf/8qcOXOYN28ePXv2ZPPmzdx5552EhobywAMPANCzZ0+WLVvmOK6X1+n/ONauXcvAgQPx9na+ifeOHTto3bo1UVFRZ9W++rSnb9++VFXV/sfk66+/JjY29rR1i4icl3wCoE03+3KymuHfo/tqDwHnH4SCg2CtgPwD9qUuJguEtD3e43fyEtJWPYDiNm7/mzdy5EhGjhzZoH1mzpzJpEmTuPPOOwF48803+eKLL3j77bd5/PHHAfjuu++47rrruPrqqwFITEzk/fffd+oZ8/LyIjo6ul6fabPZmDx5Mp07d2bBggVYLPYrtu/evZthw4Yxbdo0Hn300bNqX33as23btnrVKSIi9XDi8G9cHZM/bDYozjoe8PL3n/D8wPEAWHDAvuyv6zNODoDxdQRA7zp2FDl3bg94DVVRUcGWLVt44oknHOvMZjMpKSmsW3f8+kqDBw/mn//8J7/88gtdunThhx9+YM2aNcycOdOxzZ49e4iNjcXPz49Bgwbx/PPP065duzo/12w2s3jxYi699FLuuOMO5s+fT1paGsOGDeP666+vM9y5sj2uMnv2bGbPno3VanX5sUVEPIbZDCEx9qVdcu33XRIAzafvAfRv0+jNFM/V4gJeTk4OVqu11nBoVFQUu3btcrx+/PHHKSwspFu3blgsFqxWK88++yy33347AMnJycydO5euXbty+PBhnnnmGYYMGcL27dsJDg6u87NjY2NZsWIFQ4YMYdy4caxbt46UlBTmzJnT6O05k5SUFH744QdKSkqIi4vjww8/ZNCgQbW2mzx5MpMnT6awsJDQ0NCzrltE5LzWkABYcLB2AMw/CNZy+3sFB2H/2lqH8DKZucKrFZbcOfYJIKEn9QCGxqkHUE6pxQW8+vrggw/4z3/+w3vvvUfPnj3Ztm0bDz74ILGxsYwfP95p2LRPnz4kJyeTkJDABx98wF133XXK47Zr14758+dz2WWX0aFDB9566y1MzeAWOyeeSygiIm52YgDkFAGwJPvUPYD5BzFZywmozLVPAqnrDiAms32mb03YC4u3h8DQ+OrnceAT2OhNleapxQW8iIgILBYLWVlZTuuzsrKczqd75JFHePzxx7ntttsA6N27N/v37+f5559n/PjxtY4bFhZGly5d2Lt372k/Pysri3vuuYfRo0ezadMmHnroIV5//fVGb4+IiHgQsxmCo+1L/MDa79tsVOYfYt2ShQzuEVd9GZgTev/yD9h7AAvT7cup+IdXh7/qEBga7xwGAyN1H2AP1eICno+PD/369WP58uVcf/31gH0CxPLly5kyZYpju9LSUsxm56vAWCwWbLa6b2tTXFxMamoqv/nNb0752Tk5OQwfPpzu3bvz4Ycf8ssvvzB06FB8fX15+eWXG7U9IiJyHqkOgEcDO2P0HAUnXb3B3gN45HjvX8FBKEivngGcbn9dXghlefYl88e6P8fLr/o8wDp6/0Lj7e95+TR+e8Xl3B7wiouLnXrN0tLS2LZtG+Hh4bRr14433niDTz75hOXLlzu2mTZtGuPHj6d///4MHDiQWbNmUVJS4piFCjB69GieffZZ2rVrR8+ePfn++++ZOXMmEydOBGD69OmMHj2ahIQEMjIyePrpp7FYLIwdO7bOOm02GyNHjiQhIYGFCxfi5eVFjx49WLp0KcOGDaNt27Y89NBDDW5ffdsjIiLiYDZDcJR9iR9Q9zbHCpwDX8FB59dFmVB1DPJS7UudTPZeRqfevxN6A8PiwU/nczdHbg94mzdv5vLLL3e8njZtGgDjx49n7ty55OTkkJrq/Bfv1ltv5ciRIzz11FNkZmbSt29flixZ4jRR4fXXX+ePf/wj999/P9nZ2cTGxvLb3/6Wp556CoD09HTGjh1Lbm4ukZGRXHLJJaxfv57IyMg66zSbzTz33HMMGTIEH5/j/5tJSkpi2bJlp9zvTO2rb3tEREQaxC8UokMhulfd71dVQOGhEwJgevWkkBNeVx2zXwy66DCkb6r7OL4hJw39xh3vDQyNswdEs6Xx2il1cnvAGzp0qOPuEnWZMWMGM2bMqLV+ypQppx3CDA4OZtasWcyaNavO9xcsWNDQUrniiivqXH/BBReccp8zta/GmdojIiLiUl4+EN7evtTFMKAk53jvn2MI+ITXpbn2oeDsn+1LXUwW+y3hQtvah3xD20JIHIS2xRQQhW9lgf2zxKXcHvBERESkGTKZICjSvrS9sO5tKkqg4FD19f5OOgcw/yAUZYCt6pSTQbyAqwBj58MQElsd/OJOCINxx0OhX5gmhDSAAp6IiIicHZ9AiOxiX+pis0Jx9vGh4MJD9kBYmA4FhzAK0qE4C5O1ovp2cftO/VnegXX2AjoFQd+gxmhli6SAJyIiIo3DbDl+PcC4/rXerqqs5Msv/svISy7AuySrOgSmV4fAE0JhaS5UlkDOL/blVPxCTwp+JwXBkLbg7deIDW4+FPBERETEbQyTl31CRkSHU29UWQaFGXX2AjpelxfYZw4fKzj1+YAAAREnBb9Ye/ALjrE/D44BnwDXN7SJKeCJiIhI8+btD6072pdTKS+qI/ilO4fCqjIozbEvh3849bH8wo6HvZAY+x1DQmJPWBcLAa2b9TmBCngiIiLS8vkGQ5tu9qUuhgFlR08IfNWPhYftj0WH7c8rS+BYvn3J3nHqz7P42i8Bc2LoO/F5cIx9wT0hUAFPREREPJ/JBAHh9iWmT93bGIb9si+FGfalJvQVZTivKzliv1Vc/n77ciqDH4DLn2qc9pyBAp6IiIgI2EOgX6h9adP91NtVVUBx5klB8KTnRYft5/a5iQKeiIiISEN4+dhv2RbW7tTbGIb9GoC2pivrRGb3fKyIiIiIBzOZwOLtto9XwBMRERHxMAp4IiIiIh5GAU9ERETEwyjgiYiIiHgYBTwRERERD6OAJyIiIuJhFPBEREREPIwCnoiIiIiHUcATERER8TAKeCIiIiIeRgFPRERExMMo4ImIiIh4GAU8EREREQ/j5e4CxD0MwwCgsLDQ5ceurKyktLSUwsJCvL29XX58d/P09oHnt9HT2wdqoyfw9PaB2ng2an5v1/wePxUFvPNUUVERAPHx8W6uRERERBqqqKiI0NDQU75vMs4UAcUj2Ww2MjIyCA4OxmQyufTYhYWFxMfHc/DgQUJCQlx67ObA09sHnt9GT28fqI2ewNPbB2rj2TAMg6KiImJjYzGbT32mnXrwzlNms5m4uLhG/YyQkBCP/YEFz28feH4bPb19oDZ6Ak9vH6iNDXW6nrsammQhIiIi4mEU8EREREQ8jAKeuJyvry9PP/00vr6+7i6lUXh6+8Dz2+jp7QO10RN4evtAbWxMmmQhIiIi4mHUgyciIiLiYRTwRERERDyMAp6IiIiIh1HAExEREfEwCnjiUrNnzyYxMRE/Pz+Sk5PZuHGju0s6K88//zwDBgwgODiYNm3acP3117N7926nbYYOHYrJZHJa7r33XjdV3HAzZsyoVX+3bt0c7x87dozJkyfTunVrgoKCuPHGG8nKynJjxQ2XmJhYq40mk4nJkycDLfM7XL16NaNHjyY2NhaTycSnn37q9L5hGDz11FPExMTg7+9PSkoKe/bscdomLy+P22+/nZCQEMLCwrjrrrsoLi5uwlac2unaV1lZyWOPPUbv3r0JDAwkNjaWO+64g4yMDKdj1PW9v/DCC03cklM703c4YcKEWvVfddVVTtu01O8QqPNn0mQy8dJLLzm2ae7fYX1+R9Tn39ADBw5w9dVXExAQQJs2bXjkkUeoqqpySY0KeOIyCxcuZNq0aTz99NNs3bqVpKQkRowYQXZ2trtLa7BVq1YxefJk1q9fz9KlS6msrOTKK6+kpKTEabtJkyZx+PBhx/Liiy+6qeKz07NnT6f616xZ43jvoYce4n//+x8ffvghq1atIiMjgxtuuMGN1Tbcpk2bnNq3dOlSAG6++WbHNi3tOywpKSEpKYnZs2fX+f6LL77Ia6+9xptvvsmGDRsIDAxkxIgRHDt2zLHN7bffzs8//8zSpUv5/PPPWb16Nffcc09TNeG0Tte+0tJStm7dyh//+Ee2bt3KokWL2L17N9dee22tbf/0pz85fa+/+93vmqL8ejnTdwhw1VVXOdX//vvvO73fUr9DwKldhw8f5u2338ZkMnHjjTc6bdecv8P6/I4407+hVquVq6++moqKCr777jvmzZvH3Llzeeqpp1xTpCHiIgMHDjQmT57seG21Wo3Y2Fjj+eefd2NVrpGdnW0AxqpVqxzrLrvsMmPq1KnuK+ocPf3000ZSUlKd7+Xn5xve3t7Ghx9+6Fi3c+dOAzDWrVvXRBW63tSpU42OHTsaNpvNMIyW/x0CxieffOJ4bbPZjOjoaOOll15yrMvPzzd8fX2N999/3zAMw9ixY4cBGJs2bXJs8+WXXxomk8k4dOhQk9VeHye3ry4bN240AGP//v2OdQkJCcYrr7zSuMW5SF1tHD9+vHHdddedch9P+w6vu+46Y9iwYU7rWtJ3aBi1f0fU59/QxYsXG2az2cjMzHRsM2fOHCMkJMQoLy8/55rUgycuUVFRwZYtW0hJSXGsM5vNpKSksG7dOjdW5hoFBQUAhIeHO63/z3/+Q0REBL169eKJJ56gtLTUHeWdtT179hAbG0uHDh24/fbbOXDgAABbtmyhsrLS6fvs1q0b7dq1a7HfZ0VFBe+++y4TJ07EZDI51rf07/BEaWlpZGZmOn1voaGhJCcnO763devWERYWRv/+/R3bpKSkYDab2bBhQ5PXfK4KCgowmUyEhYU5rX/hhRdo3bo1F1xwAS+99JLLhr2aysqVK2nTpg1du3blvvvuIzc31/GeJ32HWVlZfPHFF9x111213mtJ3+HJvyPq82/ounXr6N27N1FRUY5tRowYQWFhIT///PM51+R1zkcQAXJycrBarU5/UQGioqLYtWuXm6pyDZvNxoMPPsjFF19Mr169HOvHjRtHQkICsbGx/Pjjjzz22GPs3r2bRYsWubHa+ktOTmbu3Ll07dqVw4cP88wzzzBkyBC2b99OZmYmPj4+tX5pRkVFkZmZ6Z6Cz9Gnn35Kfn4+EyZMcKxr6d/hyWq+m7p+Dmvey8zMpE2bNk7ve3l5ER4e3uK+22PHjvHYY48xduxYp5u4P/DAA1x44YWEh4fz3Xff8cQTT3D48GFmzpzpxmrr76qrruKGG26gffv2pKam8vvf/56RI0eybt06LBaLR32H8+bNIzg4uNbpHy3pO6zrd0R9/g3NzMys82e15r1zpYAncgaTJ09m+/btTuenAU7nu/Tu3ZuYmBiGDx9OamoqHTt2bOoyG2zkyJGO53369CE5OZmEhAQ++OAD/P393VhZ43jrrbcYOXIksbGxjnUt/Ts8n1VWVnLLLbdgGAZz5sxxem/atGmO53369MHHx4ff/va3PP/88y3illi33Xab43nv3r3p06cPHTt2ZOXKlQwfPtyNlbne22+/ze23346fn5/T+pb0HZ7qd4S7aYhWXCIiIgKLxVJrhlBWVhbR0dFuqurcTZkyhc8//5xvvvmGuLi4026bnJwMwN69e5uiNJcLCwujS5cu7N27l+joaCoqKsjPz3fapqV+n/v372fZsmXcfffdp92upX+HNd/N6X4Oo6Oja018qqqqIi8vr8V8tzXhbv/+/SxdutSp964uycnJVFVVsW/fvqYp0MU6dOhARESE4++lJ3yHAN9++y27d+8+488lNN/v8FS/I+rzb2h0dHSdP6s1750rBTxxCR8fH/r168fy5csd62w2G8uXL2fQoEFurOzsGIbBlClT+OSTT1ixYgXt27c/4z7btm0DICYmppGraxzFxcWkpqYSExNDv3798Pb2dvo+d+/ezYEDB1rk9/nOO+/Qpk0brr766tNu19K/w/bt2xMdHe30vRUWFrJhwwbH9zZo0CDy8/PZsmWLY5sVK1Zgs9kcAbc5qwl3e/bsYdmyZbRu3fqM+2zbtg2z2VxrWLOlSE9PJzc31/H3sqV/hzXeeust+vXrR1JS0hm3bW7f4Zl+R9Tn39BBgwbx008/OYX1mv+w9OjRwyVFirjEggULDF9fX2Pu3LnGjh07jHvuuccICwtzmiHUUtx3331GaGiosXLlSuPw4cOOpbS01DAMw9i7d6/xpz/9ydi8ebORlpZmfPbZZ0aHDh2MSy+91M2V19/DDz9srFy50khLSzPWrl1rpKSkGBEREUZ2drZhGIZx7733Gu3atTNWrFhhbN682Rg0aJAxaNAgN1fdcFar1WjXrp3x2GOPOa1vqd9hUVGR8f333xvff/+9ARgzZ840vv/+e8cs0hdeeMEICwszPvvsM+PHH380rrvuOqN9+/ZGWVmZ4xhXXXWVccEFFxgbNmww1qxZY3Tu3NkYO3asu5rk5HTtq6ioMK699lojLi7O2LZtm9PPZs2sw++++8545ZVXjG3bthmpqanGu+++a0RGRhp33HGHm1t23OnaWFRUZEyfPt1Yt26dkZaWZixbtsy48MILjc6dOxvHjh1zHKOlfoc1CgoKjICAAGPOnDm19m8J3+GZfkcYxpn/Da2qqjJ69eplXHnllca2bduMJUuWGJGRkcYTTzzhkhoV8MSlXn/9daNdu3aGj4+PMXDgQGP9+vXuLumsAHUu77zzjmEYhnHgwAHj0ksvNcLDww1fX1+jU6dOxiOPPGIUFBS4t/AGuPXWW42YmBjDx8fHaNu2rXHrrbcae/fudbxfVlZm3H///UarVq2MgIAAY8yYMcbhw4fdWPHZ+eqrrwzA2L17t9P6lvodfvPNN3X+3Rw/frxhGPZLpfzxj380oqKiDF9fX2P48OG12p6bm2uMHTvWCAoKMkJCQow777zTKCoqckNrajtd+9LS0k75s/nNN98YhmEYW7ZsMZKTk43Q0FDDz8/P6N69u/Hcc885hSN3O10bS0tLjSuvvNKIjIw0vL29jYSEBGPSpEm1/qPcUr/DGv/4xz8Mf39/Iz8/v9b+LeE7PNPvCMOo37+h+/btM0aOHGn4+/sbERERxsMPP2xUVla6pEZTdaEiIiIi4iF0Dp6IiIiIh1HAExEREfEwCngiIiIiHkYBT0RERMTDKOCJiIiIeBgFPBEREREPo4AnIiIi4mEU8EREzlOJiYnMmjXL3WWISCNQwBMRaQITJkzg+uuvB2Do0KE8+OCDTfbZc+fOJSwsrNb6TZs2cc899zRZHSLSdLzcXYCIiJydiooKfHx8znr/yMhIF1YjIs2JevBERJrQhAkTWLVqFa+++iomkwmTycS+ffsA2L59OyNHjiQoKIioqCh+85vfkJOT49h36NChTJkyhQcffJCIiAhGjBgBwMyZM+nduzeBgYHEx8dz//33U1xcDMDKlSu58847KSgocHzejBkzgNpDtAcOHOC6664jKCiIkJAQbrnlFrKyshzvz5gxg759+zJ//nwSExMJDQ3ltttuo6ioqHH/0ESkwRTwRESa0KuvvsqgQYOYNGkShw8f5vDhw8THx5Ofn8+wYcO44IIL2Lx5M0uWLCErK4tbbrnFaf958+bh4+PD2rVrefPNNwEwm8289tpr/Pzzz8ybN48VK1bw6KOPAjB48GBmzZpFSEiI4/OmT59eqy6bzcZ1111HXl4eq1atYunSpfz666/ceuutTtulpqby6aef8vnnn/P555+zatUqXnjhhUb60xKRs6UhWhGRJhQaGoqPjw8BAQFER0c71r/xxhtccMEFPPfcc451b7/9NvHx8fzyyy906dIFgM6dO/Piiy86HfPE8/kSExP5y1/+wr333svf//53fHx8CA0NxWQyOX3eyZYvX85PP/1EWloa8fHxAPz73/+mZ8+ebNq0iQEDBgD2IDh37lyCg4MB+M1vfsPy5ct59tlnz+0PRkRcSj14IiLNwA8//MA333xDUFCQY+nWrRtg7zWr0a9fv1r7Llu2jOHDh9O2bVuCg4P5zW9+Q25uLqWlpfX+/J07dxIfH+8IdwA9evQgLCyMnTt3OtYlJiY6wh1ATEwM2dnZDWqriDQ+9eCJiDQDxcXFjB49mr/+9a+13ouJiXE8DwwMdHpv3759XHPNNdx33308++yzhIeHs2bNGu666y4qKioICAhwaZ3e3t5Or00mEzabzaWfISLnTgFPRKSJ+fj4YLVandZdeOGFfPzxxyQmJuLlVf9/mrds2YLNZuNvf/sbZrN9UOaDDz444+edrHv37hw8eJCDBw86evF27NhBfn4+PXr0qHc9ItI8aIhWRKSJJSYmsmHDBvbt20dOTg42m43JkyeTl5fH2LFj2bRpE6mpqXz11Vfceeedpw1nnTp1orKyktdff51ff/2V+fPnOyZfnPh5xcXFLF++nJycnDqHblNSUujduze33347W7duZePGjdxxxx1cdtll9O/f3+V/BiLSuBTwRESa2PTp07FYLPTo0YPIyEgOHDhAbGwsa9euxWq1cuWVV9K7d28efPBBwsLCHD1zdUlKSmLmzJn89a9/pVevXvznP//h+eefd9pm8ODB3Hvvvdx6661ERkbWmqQB9qHWzz77jFatWnHppZeSkpJChw4dWLhwocvbLyKNz2QYhuHuIkRERETEddSDJyIiIuJhFPBEREREPIwCnoiIiIiHUcATERER8TAKeCIiIiIeRgFPRERExMMo4ImIiIh4GAU8EREREQ+jgCciIiLiYRTwRERERDyMAp6IiIiIh1HAExEREfEw/x9KjvhNIwm4yAAAAABJRU5ErkJggg==",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# vizualize mnist\n",
    "\n",
    "for item in history.history:\n",
    "    print(\"Key:\",item)\n",
    "\n",
    "plt.plot(history.history['loss'],label=\"train\")\n",
    "plt.plot(history.history['val_loss'],label=\"validation\")\n",
    "plt.title('Model Loss')\n",
    "plt.yscale('log')\n",
    "plt.ylabel('Cross Entropy')\n",
    "plt.xlabel('Iteration')\n",
    "plt.grid()\n",
    "plt.legend()\n",
    "plt.show()\n",
    "\n",
    "#print(\"history\",history.history)\n",
    "plt.plot(history.history['sparse_categorical_accuracy'],label=\"train\")\n",
    "plt.plot(history.history['val_sparse_categorical_accuracy'],label=\"validation\")\n",
    "plt.title('Model Accuracy')\n",
    "#plt.yscale('log')\n",
    "plt.ylabel('Acc')\n",
    "plt.xlabel('Iteration')\n",
    "plt.grid()\n",
    "plt.legend()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 3.\n",
    "\n",
    "Palyginkite savarankiškai adaptuotą kodą su Lab31.\n",
    "\n",
    "Mano variantas yra daug lėtis ir blogesnis, su laiko mano tikslumas mažėja."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 4.\n",
    "\n",
    "Išmėginkite keletą tinklo architektūrų ir mokymosi parametrų."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 60,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model: \"sequential_37\"\n",
      "_________________________________________________________________\n",
      " Layer (type)                Output Shape              Param #   \n",
      "=================================================================\n",
      " conv2d_77 (Conv2D)          (None, 26, 26, 28)        280       \n",
      "                                                                 \n",
      " dropout_42 (Dropout)        (None, 26, 26, 28)        0         \n",
      "                                                                 \n",
      " max_pooling2d_52 (MaxPooli  (None, 13, 13, 28)        0         \n",
      " ng2D)                                                           \n",
      "                                                                 \n",
      " conv2d_78 (Conv2D)          (None, 11, 11, 32)        8096      \n",
      "                                                                 \n",
      " dropout_43 (Dropout)        (None, 11, 11, 32)        0         \n",
      "                                                                 \n",
      " max_pooling2d_53 (MaxPooli  (None, 5, 5, 32)          0         \n",
      " ng2D)                                                           \n",
      "                                                                 \n",
      " conv2d_79 (Conv2D)          (None, 3, 3, 32)          9248      \n",
      "                                                                 \n",
      " dropout_44 (Dropout)        (None, 3, 3, 32)          0         \n",
      "                                                                 \n",
      " max_pooling2d_54 (MaxPooli  (None, 1, 1, 32)          0         \n",
      " ng2D)                                                           \n",
      "                                                                 \n",
      " flatten_36 (Flatten)        (None, 32)                0         \n",
      "                                                                 \n",
      " dense_79 (Dense)            (None, 64)                2112      \n",
      "                                                                 \n",
      " dense_80 (Dense)            (None, 10)                650       \n",
      "                                                                 \n",
      "=================================================================\n",
      "Total params: 20386 (79.63 KB)\n",
      "Trainable params: 20386 (79.63 KB)\n",
      "Non-trainable params: 0 (0.00 Byte)\n",
      "_________________________________________________________________\n",
      "Epoch 1/20\n",
      "1875/1875 [==============================] - 13s 7ms/step - loss: 1.0415 - sparse_categorical_accuracy: 0.6445 - val_loss: 1.2608 - val_sparse_categorical_accuracy: 0.9006\n",
      "Epoch 2/20\n",
      "1875/1875 [==============================] - 13s 7ms/step - loss: 0.5399 - sparse_categorical_accuracy: 0.8257 - val_loss: 1.1064 - val_sparse_categorical_accuracy: 0.9159\n",
      "Epoch 3/20\n",
      "1875/1875 [==============================] - 11s 6ms/step - loss: 0.4214 - sparse_categorical_accuracy: 0.8656 - val_loss: 0.8521 - val_sparse_categorical_accuracy: 0.9504\n",
      "Epoch 4/20\n",
      "1875/1875 [==============================] - 12s 6ms/step - loss: 0.3621 - sparse_categorical_accuracy: 0.8860 - val_loss: 0.8352 - val_sparse_categorical_accuracy: 0.9386\n",
      "Epoch 5/20\n",
      "1875/1875 [==============================] - 14s 7ms/step - loss: 0.3244 - sparse_categorical_accuracy: 0.8982 - val_loss: 0.7719 - val_sparse_categorical_accuracy: 0.9237\n",
      "Epoch 6/20\n",
      "1875/1875 [==============================] - 12s 6ms/step - loss: 0.2937 - sparse_categorical_accuracy: 0.9093 - val_loss: 0.7067 - val_sparse_categorical_accuracy: 0.9461\n",
      "Epoch 7/20\n",
      "1875/1875 [==============================] - 11s 6ms/step - loss: 0.2752 - sparse_categorical_accuracy: 0.9154 - val_loss: 0.6561 - val_sparse_categorical_accuracy: 0.9558\n",
      "Epoch 8/20\n",
      "1875/1875 [==============================] - 11s 6ms/step - loss: 0.2633 - sparse_categorical_accuracy: 0.9182 - val_loss: 0.6981 - val_sparse_categorical_accuracy: 0.9540\n",
      "Epoch 9/20\n",
      "1875/1875 [==============================] - 11s 6ms/step - loss: 0.2484 - sparse_categorical_accuracy: 0.9235 - val_loss: 0.6180 - val_sparse_categorical_accuracy: 0.9490\n",
      "Epoch 10/20\n",
      "1875/1875 [==============================] - 11s 6ms/step - loss: 0.2417 - sparse_categorical_accuracy: 0.9252 - val_loss: 0.6084 - val_sparse_categorical_accuracy: 0.9554\n",
      "Epoch 11/20\n",
      "1875/1875 [==============================] - 11s 6ms/step - loss: 0.2332 - sparse_categorical_accuracy: 0.9284 - val_loss: 0.6027 - val_sparse_categorical_accuracy: 0.9597\n",
      "Epoch 12/20\n",
      "1875/1875 [==============================] - 11s 6ms/step - loss: 0.2304 - sparse_categorical_accuracy: 0.9287 - val_loss: 0.5875 - val_sparse_categorical_accuracy: 0.9547\n",
      "Epoch 13/20\n",
      "1875/1875 [==============================] - 11s 6ms/step - loss: 0.2217 - sparse_categorical_accuracy: 0.9315 - val_loss: 0.5443 - val_sparse_categorical_accuracy: 0.9643\n",
      "Epoch 14/20\n",
      "1875/1875 [==============================] - 11s 6ms/step - loss: 0.2142 - sparse_categorical_accuracy: 0.9329 - val_loss: 0.5337 - val_sparse_categorical_accuracy: 0.9641\n",
      "Epoch 15/20\n",
      "1875/1875 [==============================] - 11s 6ms/step - loss: 0.2129 - sparse_categorical_accuracy: 0.9346 - val_loss: 0.5420 - val_sparse_categorical_accuracy: 0.9625\n",
      "Epoch 16/20\n",
      "1875/1875 [==============================] - 11s 6ms/step - loss: 0.2060 - sparse_categorical_accuracy: 0.9365 - val_loss: 0.5098 - val_sparse_categorical_accuracy: 0.9650\n",
      "Epoch 17/20\n",
      "1875/1875 [==============================] - 11s 6ms/step - loss: 0.2029 - sparse_categorical_accuracy: 0.9361 - val_loss: 0.5653 - val_sparse_categorical_accuracy: 0.9631\n",
      "Epoch 18/20\n",
      "1875/1875 [==============================] - 12s 6ms/step - loss: 0.2015 - sparse_categorical_accuracy: 0.9377 - val_loss: 0.5545 - val_sparse_categorical_accuracy: 0.9608\n",
      "Epoch 19/20\n",
      "1875/1875 [==============================] - 11s 6ms/step - loss: 0.1964 - sparse_categorical_accuracy: 0.9391 - val_loss: 0.5276 - val_sparse_categorical_accuracy: 0.9645\n",
      "Epoch 20/20\n",
      "1875/1875 [==============================] - 11s 6ms/step - loss: 0.1960 - sparse_categorical_accuracy: 0.9396 - val_loss: 0.5144 - val_sparse_categorical_accuracy: 0.9613\n"
     ]
    }
   ],
   "source": [
    "mnist = tf.keras.datasets.mnist\n",
    "(train_images0, train_labels0), (test_images0, test_labels0) = mnist.load_data()\n",
    "\n",
    "test_images=test_images0\n",
    "train_images=train_images0\n",
    "\n",
    "test_images  = test_images/255.0\n",
    "train_images = train_images/255.0\n",
    "\n",
    "keras_model = tf.keras.models.Sequential([\n",
    "    tf.keras.layers.Conv2D(28, (3, 3), activation='relu', input_shape=(28, 28, 1)),\n",
    "    tf.keras.layers.Dropout(.6, input_shape=(2,)),\n",
    "    tf.keras.layers.MaxPooling2D(2, 2),\n",
    "\n",
    "    tf.keras.layers.Conv2D(32, (3, 3), activation='relu'),\n",
    "    tf.keras.layers.Dropout(.6, input_shape=(2,)),\n",
    "    tf.keras.layers.MaxPooling2D((2, 2)),\n",
    "\n",
    "    tf.keras.layers.Conv2D(32, (3, 3), activation='relu'),\n",
    "    tf.keras.layers.Dropout(.6, input_shape=(2,)),\n",
    "    tf.keras.layers.MaxPooling2D((2, 2)),\n",
    "\n",
    "\n",
    "    tf.keras.layers.Flatten(),\n",
    "    tf.keras.layers.Dense(64, activation='relu'),\n",
    "    tf.keras.layers.Dense(10, activation='softmax')\n",
    "])\n",
    "\n",
    "keras_model.build(input_shape=[None,784])\n",
    "\n",
    "keras_model.summary()\n",
    "\n",
    "keras_model.compile(\n",
    "    optimizer=tf.keras.optimizers.SGD(0.02),\n",
    "    loss=tf.keras.losses.SparseCategoricalCrossentropy(from_logits=False),\n",
    "    metrics=[tf.keras.metrics.SparseCategoricalAccuracy()],\n",
    ")\n",
    "\n",
    "# Train loop\n",
    "history = keras_model.fit(\n",
    "    train_images,\n",
    "    train_labels0,\n",
    "    batch_size=100,\n",
    "    epochs=20,\n",
    "    validation_data=(test_images, test_labels0),\n",
    ")\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 61,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Key: loss\n",
      "Key: sparse_categorical_accuracy\n",
      "Key: val_loss\n",
      "Key: val_sparse_categorical_accuracy\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAloAAAHHCAYAAABnS/bqAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjguMywgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy/H5lhTAAAACXBIWXMAAA9hAAAPYQGoP6dpAAB4eElEQVR4nO3dd3gU5d7G8e/uppNGIKSQQOgQSkCaNOkiKM0CCiqgYgsHEbGe99iOXQ9iQdCjFDsWQI8iSkd6DUU6BGkhEEoqpOzO+8eQQEgoCUk25f5c11zJzj4785sMgZtnnnnGYhiGgYiIiIgUOauzCxAREREprxS0RERERIqJgpaIiIhIMVHQEhERESkmCloiIiIixURBS0RERKSYKGiJiIiIFBMFLREREZFioqAlIiIiUkwUtERErsBisfDiiy8W+HP79+/HYrEwbdq0Iq9JRMoGBS0RKROmTZuGxWLBYrGwbNmyPO8bhkF4eDgWi4VbbrnFCRUW3uLFi7FYLPzwww/OLkVEipiCloiUKR4eHnz99dd51i9ZsoRDhw7h7u7uhKpERPKnoCUiZUqfPn34/vvvycrKyrX+66+/pmXLlgQHBzupMhGRvBS0RKRMueuuuzhx4gTz5s3LWZeRkcEPP/zAkCFD8v1MamoqTzzxBOHh4bi7u9OgQQPeeecdDMPI1S49PZ3HH3+cwMBAfHx86NevH4cOHcp3m4cPH+a+++4jKCgId3d3GjduzJQpU4ruQPOxb98+7rjjDgICAvDy8uL666/n119/zdPugw8+oHHjxnh5eVG5cmVatWqVqxcwOTmZMWPGEBERgbu7O9WqVaNnz55s2LChWOsXqYgUtESkTImIiKBdu3Z88803Oet+++03EhMTufPOO/O0NwyDfv368e6773LTTTcxfvx4GjRowJNPPsnYsWNztX3ggQeYMGECN954I2+88Qaurq7cfPPNebYZHx/P9ddfz/z58xk1ahTvvfcedevW5f7772fChAlFfszZ+2zfvj2///47jz76KK+++ipnz56lX79+zJo1K6fdf//7X0aPHk1kZCQTJkzgpZdeonnz5qxevTqnzcMPP8ykSZO47bbb+Oijjxg3bhyenp5s3769WGoXqdAMEZEyYOrUqQZgrF271vjwww8NHx8fIy0tzTAMw7jjjjuMrl27GoZhGDVr1jRuvvnmnM/Nnj3bAIxXXnkl1/Zuv/12w2KxGHv27DEMwzBiYmIMwHj00UdztRsyZIgBGC+88ELOuvvvv98ICQkxEhIScrW98847DT8/v5y6YmNjDcCYOnXqZY9t0aJFBmB8//33l2wzZswYAzD+/PPPnHXJyclGrVq1jIiICMNutxuGYRj9+/c3GjdufNn9+fn5GdHR0ZdtIyJFQz1aIlLmDBo0iDNnzvDLL7+QnJzML7/8csnLhnPmzMFmszF69Ohc65944gkMw+C3337LaQfkaTdmzJhcrw3D4Mcff6Rv374YhkFCQkLO0qtXLxITE4vlEtycOXNo06YNHTt2zFnn7e3Ngw8+yP79+9m2bRsA/v7+HDp0iLVr115yW/7+/qxevZojR44UeZ0ikpuCloiUOYGBgfTo0YOvv/6amTNnYrfbuf322/Nt+/fffxMaGoqPj0+u9Y0aNcp5P/ur1WqlTp06udo1aNAg1+vjx49z+vRpPvnkEwIDA3MtI0aMAODYsWNFcpwXH8fFteR3HE8//TTe3t60adOGevXqER0dzfLly3N95q233mLr1q2Eh4fTpk0bXnzxRfbt21fkNYsIuDi7ABGRwhgyZAgjR47k6NGj9O7dG39//xLZr8PhAODuu+9m2LBh+bZp1qxZidSSn0aNGrFz505++eUX5s6dy48//shHH33E888/z0svvQSYPYKdOnVi1qxZ/PHHH7z99tu8+eabzJw5k969ezutdpHySD1aIlImDRw4EKvVyqpVqy552RCgZs2aHDlyhOTk5Fzrd+zYkfN+9leHw8HevXtztdu5c2eu19l3JNrtdnr06JHvUq1ataI4xDzHcXEt+R0HQKVKlRg8eDBTp07lwIED3HzzzTmD57OFhITw6KOPMnv2bGJjY6lSpQqvvvpqkdctUtEpaIlImeTt7c2kSZN48cUX6du37yXb9enTB7vdzocffphr/bvvvovFYsnpwcn++v777+dqd/FdhDabjdtuu40ff/yRrVu35tnf8ePHC3M4V9SnTx/WrFnDypUrc9alpqbyySefEBERQWRkJAAnTpzI9Tk3NzciIyMxDIPMzEzsdjuJiYm52lSrVo3Q0FDS09OLpXaRikyXDkWkzLrUpbsL9e3bl65du/LPf/6T/fv3ExUVxR9//MFPP/3EmDFjcsZkNW/enLvuuouPPvqIxMRE2rdvz4IFC9izZ0+ebb7xxhssWrSItm3bMnLkSCIjIzl58iQbNmxg/vz5nDx5slDH8+OPP+b0UF18nM888wzffPMNvXv3ZvTo0QQEBDB9+nRiY2P58ccfsVrN/zffeOONBAcH06FDB4KCgti+fTsffvghN998Mz4+Ppw+fZqwsDBuv/12oqKi8Pb2Zv78+axdu5b//Oc/hapbRC7DuTc9iohcnQund7ici6d3MAxzGoTHH3/cCA0NNVxdXY169eoZb7/9tuFwOHK1O3PmjDF69GijSpUqRqVKlYy+ffsaBw8ezDO9g2EYRnx8vBEdHW2Eh4cbrq6uRnBwsNG9e3fjk08+yWlT0OkdLrVkT+mwd+9e4/bbbzf8/f0NDw8Po02bNsYvv/ySa1sff/yxccMNNxhVqlQx3N3djTp16hhPPvmkkZiYaBiGYaSnpxtPPvmkERUVZfj4+BiVKlUyoqKijI8++uiyNYpI4VgM46KpkUVERESkSGiMloiIiEgxUdASERERKSYKWiIiIiLFREFLREREpJgoaImIiIgUEwUtERERkWKiCUudyOFwcOTIEXx8fLBYLM4uR0RERK6CYRgkJycTGhqaM1nwpShoOdGRI0cIDw93dhkiIiJSCAcPHiQsLOyybRS0nMjHxwcwT5Svr2+RbjszM5M//viDG2+8EVdX1yLddmmjYy2/KtLx6ljLr4p0vBXlWJOSkggPD8/5d/xyFLScKPtyoa+vb7EELS8vL3x9fcv1H3bQsZZnFel4dazlV0U63op0rMBVDfvRYHgRERGRYqKgJSIiIlJMFLREREREionGaImIiBQBu91OZmZmnvWZmZm4uLhw9uxZ7Ha7EyorOeXlWF1dXbHZbEWyLQUtERGRa2AYBkePHuX06dOXfD84OJiDBw+W+zkTy9Ox+vv7ExwcfM3HoaAlIiJyDbJDVrVq1fDy8srzD7PD4SAlJQVvb+8rTm5Z1pWHYzUMg7S0NI4dOwZASEjINW1PQUtERKSQ7HZ7TsiqUqVKvm0cDgcZGRl4eHiU2fBxtcrLsXp6egJw7NgxqlWrdk2XEcvuT0FERMTJssdkeXl5ObkSKWrZ5zS/cXcFoaAlIiJyjcr6eCTJq6jOqYKWiIiISDFR0BIREZFrEhERwYQJE5xdRqmkwfAiIiIVUJcuXWjevHmRBKS1a9dSqVKlay+qHFLQKqcse+ZjMbKcXYaIiJRRhmFgt9txcblyVAgMDATMuw4lN106LI/WfobLjDtpFTsR7BnOrkZEREqZ4cOHs2TJEt577z0sFgsWi4Vp06ZhsVj47bffaNmyJe7u7ixbtoy9e/fSv39/goKC8Pb2pnXr1syfPz/X9i6+dGiz2fj0008ZOHAgXl5e1KtXj59//rmEj7J0UNAqj/xrYNjcCU1cj23mA5ClsCUiUlIMwyAtIyvXcibDnmddcSyGYVxVje+99x7t2rVj5MiRxMXFERcXR3h4OADPPPMMb7zxBtu3b6dZs2akpKTQp08fFixYwMaNG7npppvo27cvBw4cuOw+XnrpJQYNGsTmzZvp06cPQ4cO5eTJk9f88y1rdOmwPKrXE/sdn2OZcTe2XXPg+2Fwx3RwcXN2ZSIi5d6ZTDuRz//ulH1ve7kXXm5X/qfdz88PNzc3vLy8CA4OBmDHjh0AvPzyy/Ts2TOnbUBAAFFRUTmv//3vfzNr1ix+/vlnRo0adcl9DB8+nLvuuguA1157jffff581a9Zw0003FerYyir1aJVTRp3urK49BsPFA3bOge/uhax0Z5clIiKlXKtWrXK9TklJYdy4cTRq1Ah/f3+8vb3Zvn37FXu0mjVrlvN9pUqV8PX1zXmsTUWiHq1y7LhvU+x3fInL93fDrt/MsDXoc3Bxd3ZpIiLllqerjW0v98p57XA4SE5KxsfXp9gfS+PpWvhHxWS7+O7BcePGMW/ePN555x3q1q2Lp6cnt99+OxkZlx+W4urqmuu1xWKpkIPlFbTKOaN2F7jrW/jmTtg1F2bcDYO+AFcPZ5cmIlIuWSyWXJfvHA4HWW42vNxcStXz/9zc3LDb7Vdst3z5coYPH87AgQMBs4dr//79xVxd+VF6zrgUnzpdYcgMcPGE3X+YYSvzrLOrEhERJ4qIiGD16tXs37+fhISES/Y21atXj5kzZxITE8OmTZsYMmRIheyZKiwFrYqidhcY+p0ZtvbMgxlDFbZERCqwcePGYbPZiIyMJDAw8JJjrsaPH0/lypVp3749ffv2pVevXlx33XUlXG3ZpUuHFUmtG2Do9/D1INgzH769C+78Glw9nV2ZiIiUsPr167Ny5cpc64YPH56nXUREBAsXLsy1Ljo6Otfr7EuJ2T1ddrs9z2XS06dPX1vBZZR6tCqaWp3MsOXqBXsXwjd3QeYZZ1clIiJSLiloVUQRHWHoD+BaCfYtMgfKZ6Q5uyoREZFyR0GrooroAHf/CG7esG+xwpaIiEgxUNCqyGq2Ox+2YpeYY7cyUp1dlYiISLmhoFXR1bge7p4Jbj6w/0/4erDCloiISBFR0BKo0RbuuSBsfaWeLRERkaKgoCWm8DZwzyxw94W/l8FXd0B6irOrEhERKdMUtOS88NYXhK3l8NXtkJ7s7KpERETKLAUtyS2sFdwzG9z94MBK+FJhS0REpLAUtCSvsJZw72zw8IODq+DL2+BskrOrEhGRUiQiIoIJEybkvLZYLMyePfuS7ffv34/FYiEmJuaa9ltU2ykpClqSv+rXwb0/gYc/HFytsCUiIpcVFxdH7969i3Sbw4cPZ8CAAbnWhYeHExcXR5MmTYp0X8VFQUsuLbTF+bB1aA18eSucTXR2VSIiUgoFBwfj7u5e7Pux2WwEBwfj4lI2HtesoCWXF9ochv0MnpXh0Fr4QmFLRKSs++STTwgNDc15CHS2/v37c99997F371769+9PUFAQ3t7etG7dmvnz5192mxdfOlyzZg0tWrTAw8ODVq1asXHjxlzt7XY7999/P7Vq1cLT05MGDRrw3nvv5bz/4osvMn36dH766ScsFgsWi4XFixfne+lwyZIltGnTBnd3d0JCQnjmmWfIysrKeb9Lly6MHj2ap556ioCAAIKDg3nxxRcL/oMrBAUtubKQKLj3XNg6vA6+GAhnTju7KhGR0skwzLkIL1wy0/KuK47FMK6qxDvuuIMTJ06waNGinHUnT55k7ty5DB06lJSUFPr06cOCBQvYuHEjN910E3379uXAgQNXtf2UlBRuueUWIiMjWb9+PS+++CLjxo3L1cbhcBAWFsb333/Ptm3beP7553nuuef47rvvABg3bhyDBg3ipptuIi4ujri4ONq3b59nX4cPH6ZPnz60bt2aTZs2MWnSJD777DNeeeWVXO2mT59OpUqVWL16NW+99RYvv/wy8+bNu6rjuRZlo99NnC+kGQz7H0zvB4fXwxcDzPDl4evsykRESpfMNHgtNOelFfAvqX0/dwTcKl2xWeXKlenduzdff/013bt3B+CHH36gatWqdO3aFavVSlRUVE77f//738yaNYuff/6ZUaNGXXH7X3/9NQ6Hg88++wwPDw8aN27MoUOHeOSRR3LauLq68tJLL+W8rlWrFitXruS7775j0KBBeHt74+npSXp6OsHBwZfc10cffUR4eDgffvghFouFhg0bcuTIEZ5++mmef/55rFazT6lZs2a88MILANSrV48PP/yQBQsW0LNnzysez7VQj5ZcveCmZtjyqgJHNsKSN51dkYiIFNLQoUP58ccfSU9PB+Crr77izjvvxGq1kpKSwrhx42jUqBH+/v54e3uzffv2q+7R2rFjB82aNcPDwyNnXbt27fK0mzhxIi1btiQwMBBvb28++eSTq95Htu3bt9OuXTssFkvOug4dOpCSksKhQ4dy1jVr1izX50JCQjh27FiB9lUY6tGSggluAgMmw9d3wNrPoP1o8AlydlUiIqWHq5fZs3SOw+EgKTkZXx+fnN6VYt33Verbty+GYfDrr7/SunVr/vzzT959913AvGw3b9483nnnHerWrYunpye33347GRkZRVbqt99+y7hx4/jPf/5Du3bt8PHx4e2332b16tVFto8Lubq65nptsVjyjFErDgpaUnD1ekL1VuZ4reXvwU2vObsiEZHSw2LJffnO4QBXu7muuINWAXh4eHDrrbfy1VdfsWfPHho0aMB1110HwPLlyxk+fDgDBw4EzDFX+/fvv+ptN2zYkC+//JKzZ8/m9GqtWrUqV5vly5fTvn17Hn300Zx1e/fuzdXGzc0Nu91+2X01atSIH3/8EcMwcnq1li9fjo+PD2FhYVddc3EpPWdcyg6LBbo+a36/7jNIPurcekREpFCGDh3Kr7/+ypQpUxg6dGjO+nr16jFz5kxiYmLYtGkTQ4YMKVDvz5AhQ7BYLIwcOZJt27YxZ84c3nnnnVxt6tWrx7p16/j999/ZtWsX//rXv1i7dm2uNhEREWzevJmdO3eSkJBAZmZmnn09+uijHDx4kH/84x/s2LGDn376iRdeeIGxY8cWfw/iVXB+BVI21ekOYW0g66zZqyUiImVOt27dCAgIYOfOnQwZMiRn/fjx46lcuTLt27enb9++9OrVK6e362p4e3vzv//9jy1bttCiRQv++c9/8uabucf1PvTQQ9x6660MHjyYtm3bcuLEiVy9WwAjR46kQYMGtGrVisDAQJYvX55nX9WrV2fOnDmsWbOGqKgoHn74Ye6//37+7//+r4A/jeKhS4dSONm9Wl8MhHVToMNj4HPpu0JERKT0sVqtHDlyJM/6iIgIFi5cmGtddHR0rtcXX0o0DMMcj5ZkPkXk+uuvz/OYHOOC6Sfc3d2ZOnUqU6dOzdXm9ddfz/k+MDCQP/74I099xkXTWHTu3Jk1a9bkaZdt8eLFedZd7nFBRUk9WlJ4tbtCeFuzV2vZu86uRkREpNRR0JLCs1igS/ZYramQlPd/RSIiIhWZgpZcm9pdoEY7sKerV0tEROQiClpybS7s1Vo/DRIPO7UcERGR0kRBqxxasSeBp2ZuZWmc5cqNi0KtG6BmB7BnwLLxJbNPEZFS5OLB2VL2FdU5VdAqh2JPpDJr4xH+OlVCQevCXq0Nn0Piocu3FxEpJ7JnG09LS3NyJVLUss/pxTPKF5SmdyiHmlb3A+BAqqXk/pdVqxNEdIL9f8Kf/4FbNF5LRMo/m82Gv79/zjPzvLy8cj1zD8xH8GRkZHD27NlSMYFmcSoPx2oYBmlpaRw7dgx/f39sNts1bU9B6xr98ssvPPHEEzgcDp5++mkeeOABZ5dEg2AfXG0W0rLg8Omz1KrmVjI77vIsTPsTNnwBHR8H/xols18REScKDjbnELzUA4oNw+DMmTN4enrmCWHlTXk6Vn9//5xzey0UtK5BVlYWY8eOZdGiRfj5+dGyZUsGDhxIlSpVnFqXu4uNetW82RaXzNYjSdSq5lsyO47oYI7Xil0Kf46HvhNKZr8iIk5ksVgICQmhWrVq+T4iJjMzk6VLl3LDDTdc82Wo0q68HKurq+s192RlU9C6BmvWrKFx48ZUr14dgN69e/PHH39w1113ObkyaBLqy7a4ZP46kkTf5iW44y7PmkFr45fQaax6tUSkwrDZbPn+42yz2cjKysLDw6NMh4+rUZGO9WqVzQuoRWTp0qX07duX0NBQLBZLvtPxT5w4kYiICDw8PGjbtm2uKf6PHDmSE7LAfN7S4cOlY3qDJtXNXqytR5JKdsc120OtzuDIhKXvXLm9iIhIOVahg1ZqaipRUVFMnDgx3/dnzJjB2LFjeeGFF9iwYQNRUVH06tXrktfhS5MmoeeC1uGkkr/tuOtz5teYr+DU/pLdt4iISClSoS8d9u7dm969e1/y/fHjxzNy5EhGjBgBwOTJk/n111+ZMmUKzzzzDKGhobl6sA4fPkybNm0uub309HTS09NzXmc/eDMzMzPf6/rXonaABzaLwekzmew/nkxYZc8i3f5lhbTEVqsL1tjFOBa/hf2W94p1d9k/u6L+GZZGFelYoWIdr461/KpIx1tRjrUgx2cxNMsaYA5mnDVrFgMGDAAgIyMDLy8vfvjhh5x1AMOGDeP06dP89NNPZGVl0ahRIxYvXpwzGH7FihWXHAz/4osv8tJLL+VZ//XXX+Pl5VXkx/T2ZhuHUi2MqG+neZWSPc2VU3dzw65/48DKgsi3SHOvVqL7FxERKS5paWkMGTKExMREfH0vf8NZhe7RupyEhATsdjtBQUG51gcFBbFjxw4AXFxc+M9//kPXrl1xOBw89dRTl73j8Nlnn2Xs2LE5r5OSkggPD+fGG2+84okqqMzMTL7du4BDqRbcg+vSp2e9It3+1XB8sxzrvoV0d9mAvc/7xbafzMxM5s2bR8+ePcv94MuKdKxQsY5Xx1p+VaTjrSjHmn1F6mooaF2jfv360a9fv6tq6+7ujru7e571rq6uxfIHMtzbYOUx+Csu2Tl/4Lv9E/YtxLplBtbO46BKnWLdXXH9HEujinSsULGOV8daflWk4y3vx1qQY6vQg+Evp2rVqthsNuLj43Otj4+PL5IJzEpCeCXzcuGWw4nOeQ5XWCuo2xMMu+5AFBGRCklB6xLc3Nxo2bIlCxYsyFnncDhYsGAB7dq1c2JlVy/UC1xtFk6nZXLo1BnnFJH9DMTN38KJvc6pQURExEkqdNBKSUkhJiaGmJgYAGJjY4mJieHAgQMAjB07lv/+979Mnz6d7du388gjj5CamppzF2Jp52KF+kHeAGw9nOicIsJaQr1eYDhg6dvOqUFERMRJKnTQWrduHS1atKBFixaAGaxatGjB888/D8DgwYN55513eP7552nevDkxMTHMnTs3zwD50ix7Pq0tzgpaAF2eMb9ungEJe5xXh4iISAmr0IPhu3TpcsWxS6NGjWLUqFElVFHRaxzqCxx2btCqfh3U7w27foOlb8GtnzivFhERkRJUoXu0KoILe7ScOmVal6fNr1u+h4TdzqtDRESkBClolXP1g3ycPyAeILQFNOhjjtVa8qbz6hARESlBClrlnLuLlQbBPoATB8Rnyx6rteUHOL7TubWIiIiUAAWtCqBpdT8ANjs7aIVEQcNbAEO9WiIiUiEoaFUATc4FLaf3aMH5Xq2tM+HYdufWIiIiUswUtCqA7B4tpw+IBwhuCo36ol4tERGpCBS0KoAGwaVkQHy2zud6tf6aDfHbnFqKiIhIcVLQqgDcXWw5A+KdOp9WtuAmENkf9WqJiEh5p6BVQVx4+bBU6PwMYIFtsyH+L2dXIyIiUiwUtCqIUjUgHiAoEhoPML9f/IZTSxERESkuCloVRLPq/kApGRCfrfPTgAW2/wxHtzi7GhERkSKnoOUEEydOJDIyktatW5fYPusHe5euAfEA1RpB44Hm9+rVEhGRckhBywmio6PZtm0ba9euLbF9lroB8dmye7V2/AJxm51djYiISJFS0KpASt2AeIBqDaHJbeb36tUSEZFyRkGrAil1A+KzdX4aLFbY+SsciXF2NSIiIkVGQasCKZUD4gEC60OT283v1aslIiLliIJWBVIqB8Rny+7V2vUbHN7g7GpERESKhIJWBVJqB8QDVK0LTQeZ32u2eBERKScUtCqYUjkgPlvnp8Big11z4fB6Z1cjIiJyzRS0Kpim58ZplboB8QBV6kCzweb3816Ag2vhzCnn1iQiInINXJxdgJSsC3u0DMPAYrE4uaKL3DAONs+A/X/CZz3MdV5VoWo9qFIXqtY/9309qFwTbK7OrVdEROQyFLQqmIsHxIcHeDm7pNyq1IG+78GW7yBhNyTHQVoCHEiAAytzt7W6QOVa2KrUJfK0FUvMKQhqaIawSlWcU7+IiMgFFLQqmOwB8VsPJ7HlcGLpC1oA191jLgDpyXBiDyTsgRO7zfB1Yrf5OusMnNiN9cRu6gH8+uv5bXhWNnu/qtQzB9pXqWf2hFWuBS5uzjgqERGpgBS0KqCm1f1zglafpiHOLufy3H0gtIW5XMjhgKTDcGI39vid/L1hPhE+WVhP7oXEg+bYroOrzeVCFhtUjoAGvc3B9x5+JXYoIiJS8ShoVUBNq/vxDaV0QPzVslrBPxz8w3HU6MSW46GE9+mD1dUVMlLhxN7zPV85PWF7ICMFTu6FlR/C5u/gxn+bA/BL21g1EREpFxS0KqDsAfGbD5XSAfHXyq0ShDQzlwsZhjnm6/B6mP+SGcBmPQTrp0GfdyC4iVPKFRGR8kvTO1RA2QPiE8+Uwhnii5PFAr6h0KgvPLICerwIrl7mIPuPb4DfnoYzp51dpYiIlCMKWhVQqZ4hvqS4uEHHx2HUWogcAIYdVk+GD1tBzDdm75eIiMg1UtCqoJpe8IDpCs0vDAZNh3tmm3cmph6H2Q/DlJvg6BZnVyciImWcglYFlTNx6aEKHrSy1el67nLiS+BaCQ6uMi8nznlKlxNFRKTQFLQqqItniBfOXU4cY15ObDwQDAes+fjc5cSvzSklRERECkBBq4KqsAPir4Zfdbhjmnk5sWr9c5cTH4GpvXU5UURECkRBywkmTpxIZGQkrVu3dloN7i42Ggb7AhqndUl1usLDy3U5UURECk1Bywmio6PZtm0ba9eudWodTS64fCiXoMuJIiJyDRS0KjANiC+A7MuJ9/500eXEmyBus7OrExGRUkpBqwLTgPhCqN3FvJzY8+VzlxNXwyedYc6TupwoIiJ5KGhVYPWDvXGzWTUgvqBc3KDDY+cuJ9567nLiJ/BBS9j4lS4niohIDgWtCkwzxF8jv+pwx9RzlxMbQFoC/PQoTOkF8X85uzoRESkFFLQquCYXPGBaCql2F3h42fnLiYfWwMedYek7YM9ydnUiIuJECloVXPY4ra3q0bo22ZcT/7EOGvQBRyYs/Dd81gOO7XB2dSIi4iQKWhWcBsQXMd9QuPNrGPgJePjBkY3wcSdY9q56t0REKiAFrQpOA+KLgcUCUYPh0dVQrxfYM2D+i+bYreO7nF2diIiUIAWtCk4D4ouRbwgMmQH9PwJ3Xzi8DiZ3hBUfgMPu7OpERKQEKGiJBsQXJ4sFWgyFR1dBne5gT4c//s98bmLCHmdXJyIixUxBSzQgviT4VYe7f4S+74ObjznR6eSOsGqS5t0SESnHFLSEZmEaEF8iLBZoOQweXWlOCZF1BuY+A9NuhpP7nF2diIgUAwUtoX6QjwbElyT/cLhnNtzyrjnv1oEVMKkDrPmverdERMoZBS3BzcWaMyBe47RKiMUCre6DR1dARCfITIM54+DzfnBqv7OrExGRIqKgJcD5AfG687CEVY6Ae3+GPu+Aqxfs/xM+ag9rPwNdxhURKfMUtATQgHinslqhzUh4ZDnU7ACZqfDrWPhiAJw+4OzqRETkGihoCaAB8aVCQG0Y9gvc9Ca4eMK+xWbv1vrp6t0SESmjFLQEyD0g/uBJDYh3GqsVrn/Y7N0Kvx4ykuF/o+HL2yDxsLOrExGRAlLQEiD3gHiN0yoFqtSBEXPgxlfBxQP2LoCPrsey6Wv1bomIlCEKWpJDA+JLGasN2o+Ch/6E6q0gPQmXX0bTdt943ZkoIlJGKGhJjuxxWhoQX8oE1of7/4AeL2HY3AhO2oTLx+3hj3/BWZ0rEZHSTEFLcjStrgHxpZbVBh3HkPXAYo75NMFiz4AV78P7LWDtp2DPcnaFV5Z5Frb9BCdjnV2JiEiJUdBygokTJxIZGUnr1q2dXUouGhBfBlStz8o6T5I1+BuoWh/STsCvT8DkDrB7vrOry9/ZJFg2Ad5rBt/dCx93hgOrnF2ViEiJUNBygujoaLZt28batWudXUouGhBfRlgsGHV7wiMrzIlOPQPg+A746jbz7sRj251doSnlOCz4N0xoAvNfgJR4sLlBeiJ8PgB2z3N2hSIixU5BS3JpGqYB8WWGzdWc6HT0Rmg3CqyusGc+TGoPvzwOqQnOqev0AZjzJExoCn++Y44jq1of+n8ET+6Ber3MB2p/cyds+cE5NYqIlBAFLclFM8SXQZ7+0OtViF4NjfqC4YB1U8zxW8vfg6z0kqnj2HaY+RC81xzWfGKGqdDrYPCX8OhqaDEUPPzgzq+g6R3gyIIfHzAfpi0iUk65OLsAKV0uHhBvsVicXJFctSp1zFCzfxn8/hzEbYJ5z5vPTez5MkT2Nx9mXdQOroVl78LOX8+vq90FOj4OtTrn3afNFQZ+Ap6VzUA2ZxycOQU3PFk89YmIOJGCluRy8YD4GlW8nF2SFFRERxi5GDZ/CwtehtN/w/fDoEY7s+erestr34dhwN6FZsDa/+e5lRZodIsZsK60D6sVer9lji9b8gYsehXSTkKv18z3RETKCf2NJrloQHw5YbVC8yHwj/XQ+Rnz2YkHVsJ/u8HMBwv/OB+HHf6aBZ90hi9vNUOW1QWa3w3Ra8wetasNchYLdH3WfLYjwOpJ8NOjYM8sXG0iIqWQgpbkoQHx5YhbJTPMjN4AUXeZ6zbPgA9awsJXIT3l6raTlQ4bPoeJbeD74eZlSVcvaPsIPLYJBkw0J1YtjOsfNi8lWmyw6RuYcQ9kanoRESkfFLQkDw2IL4d8Q2HgZBi5CGq0NweqL33LDFwbvwSHI//PpafAig/NAe4//wNO7AEPf+j8NIzZCr3fAL+wa68varA5SN7FA3b9Bl/ebs6/JSJSxiloSR6aIb4cq36d+bDqQV9A5QhIOQo/RZuXAmP/PN8u7SQseg3ebQx//BOSj4BPiPmQ68e3QtfnoFKVoq2tQW+4eya4+8Lfy2D6LZB6vGj3ISJSwjQYXvLQgPhyzmKByH5Qvxes/hiWvg1HN5vBpuEt4BcOG6ZDZprZPqAOdHgMou4EF/firS2iAwz/Bb64FeI24fL5LXiGRBfvPkVEipF6tCQPNxcrDUM0IL7cc3GHDqPNCU9bP2COkdrxizkoPTMNgpvBHdNg1FpoOaz4Q1a2kCi473fwC8dyci+ddr0CCbtKZt8iIkVMQUvy1aS6BsRXGJWqws3/MR/p0/AWqNMd7v4RHloKjQeaD7QuaVXrwn2/Y1Stj2fmSVw+vwUObyj5OkRErpGCluTr/Dit084tREpOtYbmgPR7ZkLdHs6fPNSvOln3/I9TXrWxnDkJ0/vCviXOrUlEpIAUtCRf5+88TNKAeHEeryqsqPs0jogbICMFvrodtv/i7KpERK5agYNWREQEL7/8MgcOHCiOeqSUuHhAvIizZNk8sQ/+xnyOoz0DvrvHnJJCRKQMKHDQGjNmDDNnzqR27dr07NmTb7/9lvT0EnporZQYDYiXUsXFHW6fBi3uNh+a/VM0rPjA2VWJiFxRoYJWTEwMa9asoVGjRvzjH/8gJCSEUaNGsWGDBquWJ9kD4jdrnJaUBjYX6PchtP+H+fqP/4P5L5nPXRQRKaUKPUbruuuu4/333+fIkSO88MILfPrpp7Ru3ZrmzZszZcoUjespBzRDvJQ6Fgvc+Ar0eNF8vWw8/PK4+QxGEZFSqNBBKzMzk++++45+/frxxBNP0KpVKz799FNuu+02nnvuOYYOHVqUdYoTaEC8lFodH4dbJgAWWD8VfrwfsjKcXZWISB4Fnhl+w4YNTJ06lW+++Qar1cq9997Lu+++S8OGDXPaDBw4kNatWxdpoVLyNEO8lGqtRoCnP/w4Ev6aBWcTYfCX5oO0RURKiQL3aLVu3Zrdu3czadIkDh8+zDvvvJMrZAHUqlWLO++8s8iKFOfQgHgp9RoPhCEzwNUL9i6Ez/vDX7Ph0HpIOabxWyLidAXu0dq3bx81a9a8bJtKlSoxderUQhdVVg0cOJDFixfTvXt3fvjhB2eXUySaVPdj86FENh8+zc3NQpxdjkhedbvDvT+bc2wdWgvfDzv/ns0d/MLAP9x8hqNf+AXfh4FvdXBxc17tIlLuFThoZYesdevWsX37dgAaNWpEq1atirayMuixxx7jvvvuY/r06c4upchoQLyUCeGt4f4/YNm7cHIfnD4IyXFgT4eTe80lXxbwCb4ggIWd+77G+e89fEv0UESkfClw0Dp06BB33XUXy5cvx9/fH4DTp0/Tvn17vv32W8LCwoq6xjKjS5cuLF682NllFKmLB8RbnP1YFpFLCWwAAyeff52VAUmHIfEQJB40w1fiwQu+P2QGseQ4czm0Jv/tevid7w3zDQWvAPCsnP/i4a8eMhHJpcBB64EHHiAzM5Pt27fToEEDAHbu3MmIESN44IEHmDt3boGLOHz4ME8//TS//fYbaWlp1K1bl6lTpxZZL9nSpUt5++23Wb9+PXFxccyaNYsBAwbkaTdx4kTefvttjh49SlRUFB988AFt2rQpkhrKKg2IlzLLxQ0CaplLfgwDUo/nH8ASD5jfnz1tDrI/mwjxW69uv27elw5i+S3ZwU1PRBMplwoctJYsWcKKFStyQhZAgwYN+OCDD+jUqVOBCzh16hQdOnSga9eu/PbbbwQGBrJ7924qV66cb/vly5fTpk0bXF1dc63ftm0bVapUISgoKM9nUlNTiYqK4r777uPWW2/Nd7szZsxg7NixTJ48mbZt2zJhwgR69erFzp07qVatGgDNmzcnKysrz2f/+OMPQkNDC3roZUL2gPjscVoKWlJuWCzgXc1cwlrm3yY92Qxe2WEsOQ7OnIYzpy5YTppfz567vJ6RYi6JBwtUjourF91s/tgyf4bQFhASBcFNzTsrRaTMKnDQCg8PJzMzM896u91eqLDx5ptvEh4enmvwfK1a+f8P1OFwEB0dTb169fj222+x2WyA2aPWrVs3xo4dy1NPPZXnc71796Z3796XrWP8+PGMHDmSESNGADB58mR+/fVXpkyZwjPPPANATExMgY+vPMgeEL/lcCK3NCufgVIkX+4+UK2RuVyJw26GrVwh7CoXw4ElMw2fzDTY+oO5ZKtcywxdIVEQ0gxCmkOlqsV2yCJStAoctN5++23+8Y9/MHHixJxLe+vWreOxxx7jnXfeKXABP//8M7169eKOO+5gyZIlVK9enUcffZSRI0fmaWu1WpkzZw433HAD9957L1988QWxsbF069aNAQMG5BuyrkZGRgbr16/n2WefzbWvHj16sHLlykJt83ImTpzIxIkTsdvLxmzWzar78TUaEC9yWVabeRnQK6Bgn3M4ID2JzORjrJv7LW3C3bEd2wJxm+D0ATgVay7bZp//jG/1C8LXucUnxOylE5FSpcBBa/jw4aSlpdG2bVtcXMyPZ2Vl4eLiwn333cd9992X0/bkyZNX3N6+ffuYNGkSY8eO5bnnnmPt2rWMHj0aNzc3hg0blqd9aGgoCxcupFOnTgwZMoSVK1fSo0cPJk2aVNBDyZGQkIDdbs9z2TEoKIgdO3Zc9XZ69OjBpk2bSE1NJSwsjO+//5527drlaRcdHU10dDRJSUn4+fkVuu6S0kQD4kWKj9VqXh50qcQxvygcHftgyx4akXYSjm42Q1f2cmKvOcg/6TDsnHN+O5UCz11ubHY+fFWOUPgScbICB60JEyYUaQEOh4NWrVrx2muvAdCiRQu2bt3K5MmT8w1aADVq1OCLL76gc+fO1K5dm88++6xU/OM/f/58Z5dQLDQgXsRJvAKgdhdzyZaeDEe35g5fx3eYA/v3zDeXbB5+FwSv5lCvp8Z8iZSwAgetS4WfwgoJCSEyMjLXukaNGvHjjz9e8jPx8fE8+OCD9O3bl7Vr1/L444/zwQcfFLqGqlWrYrPZiI+Pz7Of4ODgQm+3vNCAeJFSxN0HarYzl2yZZyB+G8TFnA9fx7aZY8b2/2kuAJ4B0O2fcN1wsBX4r38RKYRC/abZ7XZmz56dM2Fp48aN6devX87g9ILo0KEDO3fuzLVu165dl5x9PiEhge7du9OoUSO+//57du3aRZcuXXB3dy/UGDEANzc3WrZsyYIFC3KmfXA4HCxYsIBRo0YVapvljQbEi5Rirp7mnZMX3j2ZlWH2dMVtMi8/7l0IJ/bAr0/Auqlw0xtQq+B3iotIwRQ4aO3Zs4c+ffpw+PDhnCkeXn/9dcLDw/n111+pU6dOgbb3+OOP0759e1577TUGDRrEmjVr+OSTT/jkk0/ytHU4HPTu3ZuaNWsyY8YMXFxciIyMZN68eXTr1o3q1avz+OOP5/lcSkoKe/bsyXkdGxtLTEwMAQEB1KhRA4CxY8cybNgwWrVqRZs2bZgwYQKpqak5dyFWdBoQL1LGuLidu0uxmfnangXrpsCiV805wabfApH9oee/ofLlH6smIoVX4KA1evRo6tSpw6pVqwgIMO+uOXHiBHfffTejR4/m119/LdD2WrduzaxZs3j22Wd5+eWXqVWrFhMmTGDo0KF52lqtVl577TU6deqEm9v52ZejoqKYP38+gYGB+e5j3bp1dO3aNef12LFjAfMy6LRp0wAYPHgwx48f5/nnn+fo0aM0b96cuXPn5jsvV0WUPSB+y6FEDYgXKYtsLtD2QWh6uxm21k2BbT/Brt+h/WjoOAbcKjm7SpFyp1ATll4YsgCqVKnCG2+8QYcOHQpVxC233MItt9xyVW179uyZ7/oWLVpc8jNdunTBMIwrbnvUqFG6VHgJ2QPik85mceBkGjWr6C9kkTLJKwBu/g+0HAFznzHHby19C2K+gp4vQ5PbdKeiSBEq8DMf3N3dSU5OzrM+JSUlVy+TlC/ZA+IBtujyoUjZF9wEhv0PBn0OfjXM6SJ+vB+m3ARHYpxdnUi5UeCgdcstt/Dggw+yevVqDMPAMAxWrVrFww8/TL9+/YqjRiklsh8wraAlUk5YLOY4rVFroOv/gasXHFwFn3SBn/8BKcedXaFcydlE2PWHGY7T83aCiPMV+NLh+++/z7Bhw2jXrl3O8wazsrLo168f7733XpEXKKVH05yJSxW0RMoVV0/o/CQ0HwLzX4At38OGz+Gv2dD5aWjzoDm4XkqXw+thxr2QdOj8Ou8gCKgDVbKXuubrgFrmeZYSV6CgZRgGSUlJfPvttxw+fDhneodGjRpRt27dYilQSg8NiBcp5/yqw22fQusH4LenzKkh/vgnrJ8GN71uTngqpcOGz82pOuwZ5lMBDAPSEiAl3lwOrLjoAxbwCzPDV8C5AJYdxPxrgM3VKYdRERQ4aNWtW5e//vqLevXqKVxVMBoQL1JB1LgeRi4yB8gveBlO7Iavbod6vaDXa1BVf/c7TVY6zHkSNkw3Xze4GQZOMp8CcOY0nNwLJ/aZc6ad3Gt+PbEX0pMg8aC57Fuce5sWm/m4ppwesNrng5hvmPmYKCm0AgUtq9VKvXr1OHHiBPXq1SuumqSUcnOx0ijEh03nJi5V0BIpx6w2uO5ecwzXkrdg9WTY/bs58en1D8MNT4GHr7OrrFgSD8GMe+DIBsAC3f4POo49H4Q8/aF6S3O5kGFAakLu4HViD5zcZ36fdcZ87+Re2P1H7s+6eEBwU2g22JwaxLNySRxpuVLgMVpvvPEGTz75JJMmTaJJkybFUZOUYk2q++UELc0QL1IBePhBr1eh5XD4/TnzH+IVH8Cmb6H7C9B8qHo8SkLsUvh+hHl50LOyeYm3bo+r+6zFAt6B5lLj+tzvORyQHHdBD9je80Hs1H7IOguH1prL7/+ERreY57x2FzOMyxUVOGjde++9pKWlERUVhZubG56euQfXnTx5ssiKk9JHA+JFKqiq9WDo9+Ydbr8/a/5D/PMoWPsp9H4LarR1doXlk2GYwXb+C2A4zIeED/7CvNRXFKxWc2yeX3Wo3Tn3e/YsOP037JoLG7+CY3/B1h/Nxbc6RN1l3kBRpWBPhKloChy03n33XQ2CrsA0IF6kgqt/o9mbseZj85JiXAxMudGc6LRuj3Nje+qaE6PKtUlPhp9GwbbZ5uuoIXDL+JK7e9DmYoaodtFw/aPmud74lXlXatJh+PMdc6nRHloMhcgBYHUvmdrKkAIHreHDhxdDGVJWaEC8iODiBu3/YY7bWfAybPzyfE9HNs+A86GrSh2zRyx7oHXB/+kpGvYs8zJZ0mFzvFPiQfNr0hGzxjYPgX+4c2q7WMJu+HYoJOwEqyv0fgNa3e+8WfstFghtYS43vgI755g3S+xdaN7heGAFzHkKW6N+VDlTC4zezqmzFCrwn3abzUZcXBzVqlXLtf7EiRNUq1YNu91eZMVJ6XPhgPgNB04paIlUZN7VoP+H0Pp+M2wl7DLH9yQdhjMn4dAac8nFgotvddoZ/ljnLoHA+mYAq1oX/MILP+7HMODs6XMB6tAFQerw+dfJR8zLb5eyahI0vQM6PAbVGhWujqKw/ReY9TBkJINPiDl7f3gb59VzMVcPaHKruSQdgU3fmD1dJ/di3fwNHQFj0jfmWK7md5nTSlRgBQ5al3pmYHp6uh7BU0F0rFeVTYcS+WRpLP2jqmO16vKhSIWW3dORLSPVvKMtYff5gdUn9pjTRJxNxJJ0iGocgvVbc2/H5nbB1AIXLR6+F/REHToXoA7mDlaZqVeu1epqjkfyDTMDgF+YGRh3/GIOON/0jbnUvwk6jIGa7Yr0R3VZDjssfAWWjTdf1+wAt08Fn6CSq6GgfEOh0xPm3Y8HV+NY/zmOLT/icioWFr1iPsC8dhdocTc0vLlCTpp61UHr/fffB8BisfDpp5/i7e2d857dbmfp0qU0bNiw6CuUUueBjrX5fMXfbI9L4n+bj9C/eXVnlyQipYlbJXNKgOCmudcbBqSdICt+B1sWz6RZ9UrYTp2bYuDkXnPyzeM7zKWwvKqeD1B+4ee+Vj//faVq+d8l2fYhc6b1ZRNg+//MAeC75kL49dBxjDmHWHHeXZl2En64D/YtMl9f/6j5kO+yMpGoxQI1rsce0pLfjS7cVDMDly3fmg8t37fIXNz9oOlt0PxuqH5dhXl4+VUHrXfffRcwe7QmT56MzXa+e9fNzY2IiAgmT55c9BVKqVO5khsPda7NO3/s4j9/7KJ3kxDcXHR7t4hcgcUClapihLflQJUTNOnWB9u5R7nhsJs9VBfO85TdI5Z4EDDAxfOC4HRhkDr3vW/otfWYVG9p3tGXsAdWvG/2bB1cBd/cCYGNzEuKTW8v+vBzJMacHyvxgPm8yX4fmPspo+w2d4xmA6Hl3XAy1vw5xnxtnsd1U8wlsJE5gL7ZYLNHsRy76qAVGxsLQNeuXZk5cyaVK2vSsopsRIdaTFvxNwdOpjFj3UHuub6ms0sSkbLMem528soReeeHyjxjLp6VS6YXpGpd6Pc+dH0OVn0Ea6fA8e0w+2Hz0l67aHMyV3fvK2/rSjZ+Bb88DvZ0qFwL7vwKghpf+3ZLi4Ba5s+x8zOwf6k5lm/7/8yf5x//B/NegLBWEBJlTl0R0swMYeXo2ZoFHqO1aNGi4qhDyphK7i6M7l6X53/6i/cX7Ob268LwdNPkdSJSDFw9nTO2xyfYvHzXcazZC7NqkvkA59+fhaVvmQ/bbvMQVKpS8G1nZcDcZ2DdZ+br+jfBwI/N2d3LI6vVHKtVu4v5qKC/Zpoh8/A6OLjaXHLaukK1hufCV5QZvoKaFE2wdYICBy273c60adNYsGABx44dw+HIfQfHwoULi6w4Kd3ubF2D//65j4MnzzB1RSyPdtHzz0SkHPL0h05jzXFTm74xLyue3AdL3oTl78N190C7UVD5Knv2k47Ad/eas61jgS7Pwg1PVpwZ9j39odV95nJirzk2Lm4THN0McZvNu0ePbjEXvjz3IYs5BUdwMzOAhTQzQ1hhQm4JK3DQeuyxx5g2bRo333wzTZo00YSVFZibi5WxPevz+IxNTF68l6FtauLnVUYGboqIFJSrB7QaYV423P6zOXA+LgbWfAJrPzMnbe3wGARf5vF0+5fB98Mh9bj5eKNbPzUnga2oqtQxl2aDzNeGAacPnA9d2V+Tj5y/e/Wvmec/71v9ovDVzByzV4qySYGD1rfffst3331Hnz59iqMeKWP6RVXn4yX72HE0mUlL9vJMb915KiLlnNUGjQeaM6HHLoFl78K+xbDlO3Op2xM6Pg4125//B98wsK6eBAteBMNuXgob/MW5CVwlh8Vi9gxWrgmN+p5fn3Icjm7KHb5OnpuzLekw7PrtfFvPyufHewVHQVhLp/6cCxy03NzcqFtXl4jEZLNaeLJXA+6fvo6py2MZ3j6CYD8PZ5clIlL8LJbz446ObITl78G2n2DPPHMJa23OxRXenpb7P8IWc24cUtNB0Pc9cPNyYvFljHegeZPEhTdKpCfD0a25Lzse3w5nTpkBOHaJ2a7xrXDHVOfUTSGC1hNPPMF7773Hhx9+qMuGAkC3htVoVbMy6/4+xfsLd/PawKZX/pCISHkS2gLumGaOOVrxgTmdwaG1MGMoLi4ehGWdxbC6YOn1mjmIXv9+Xjt3H3NC2Qsnlc1Kh2PbzwWvcz1g4c594HmBg9ayZctYtGgRv/32G40bN8bVNfeYnJkzZ17ik1JeWSwWnu7dkDsmr2TG2oOM7FSbWlX1aB4RqYCq1IG+E8wB7qsnw9rPsKQnctbFD5chX+FSu5OzKyzfXNwhtLm5lBIFDlr+/v4MHDiwOGqRMqx1RADdGlZj4Y5j/OePnXw45DpnlyQi4jw+QdDjBej4OFk7f2fR7rP0CL/e2VWJExQ4aE2d6rzrnFK6PdmrAYt2HuOXzXE83DmRJtX9nF2SiIhzefhiRA4gY/8cZ1ciTnLVk3YcO3bssu9nZWWxZs3FT2mXiqRRiC/9o0IBeOv3nU6uRkRExPmuOmiFhITkCltNmzbl4MGDOa9PnDhBu3Yl+JTzMmzixIlERkbSunVrZ5dS5Mb2bICL1cLSXcdZsTfB2eWIiIg41VUHLcMwcr3ev38/mZmZl20j+YuOjmbbtm2sXbvW2aUUuRpVvBjStgYAb83dqT8TIiJSoRXpfP+a7kEARnWri6erjZiDp5m3Ld7Z5YiIiDhNBXmwkpSkaj4e3NcxAoC3f9+J3aFeLRERqZiuOmhZLBaSk5NJSkoiMTERi8VCSkoKSUlJOYtItgdvqIOfpyu7j6Uwa+NhZ5cjIiLiFAUao1W/fn0qV65MQEAAKSkptGjRgsqVK1O5cmUaNGhQnHVKGePn6cqjXeoA8O68XaRn2Z1ckYiISMm76nm0Fi1aVJx1SDk0rH0EU5bHcvj0Gb5adYD7OtZydkkiIiIl6qqDVufOnYuzDimHPFxtjOlRn2dnbuHDRXsY1Docb/cCz5ErIiJSZmkwvBSrO1qGUbtqJU6mZvDpn/ucXY6IiEiJUtCSYuVis/LEjeb4vf8u3ceJlHQnVyQiIlJyFLSk2PVuEkzT6n6kZtiZuGivs8sREREpMQpaUuysVgtP3WT2an256m8OnUpzckUiIiIl45qDVlJSErNnz2b79u1FUY+UUx3rVqV9nSpk2B1MmL/b2eWIiIiUiAIHrUGDBvHhhx8CcObMGVq1asWgQYNo1qwZP/74Y5EXKOWDxWLhqZsaAjBzwyF2xyc7uSIREZHiV+CgtXTpUjp16gTArFmzMAyD06dP8/777/PKK68UeYFSfjQP96dX4yAcBrzzx05nlyMiIlLsChy0EhMTCQgIAGDu3LncdttteHl5cfPNN7N7ty4JyeWNu7EBVgv8/lc8Gw+ccnY5IiIixarAQSs8PJyVK1eSmprK3LlzufHGGwE4deoUHh4eRV6glC/1gny47bowAN6cuwPD0AOnRUSk/Cpw0BozZgxDhw4lLCyM0NBQunTpApiXFJs2bVrU9Uk5NKZnfdxsVlbtO8mfuxOcXY6IiEixKXDQevTRR1m5ciVTpkxh2bJlWK3mJmrXrq0xWnJVqvt7ck+7mgC89fsOHA71aomISPlUqOkdWrVqxcCBA/H29sZutxMTE0P79u3p0KFDUdcn5dSjXerg7e7C1sNJzNka5+xyREREikWhLh1+9tlnANjtdjp37sx1111HeHg4ixcvLur6pJyq4u3OyE61AfjPH7vItDucXJGIiEjRK3DQ+uGHH4iKigLgf//7H7GxsezYsYPHH3+cf/7zn0VeoJRf93eqRZVKbsQmpPL9ukPOLkdERKTIFThoJSQkEBwcDMCcOXO44447qF+/Pvfddx9btmwp8gKl/PJ2d2FUt7oAvLdgF2cy7E6uSEREpGgVOGgFBQWxbds27HY7c+fOpWfPngCkpaVhs9mKvEAp34a0rUF1f0/ik9KZvnK/s8sREREpUgUOWiNGjGDQoEE0adIEi8VCjx49AFi9ejUNGzYs8gKlfHN3sTG2Z30APlq0h8S0TCdXJCIiUnQKHLRefPFFPv30Ux588EGWL1+Ou7s7ADabjWeeeabIC5Tyb0CL6tQP8ibpbBYfL93r7HJERESKjEthPnT77bfnWTds2LBrLkYqJpvVwrgbG/DgF+uZunw/w9tHUM1XTxkQEZGyr1DzaC1ZsoS+fftSt25d6tatS79+/fjzzz+LujapQHpGBtGihj9nMu18sHCPs8sREREpEgUOWl9++SU9evTAy8uL0aNHM3r0aDw9PenevTtff/11cdQoFYDFYuHpm8wxft+sOcDfJ1KdXJGIiMi1K3DQevXVV3nrrbeYMWNGTtCaMWMGb7zxBv/+97+Lo0apIK6vXYXO9QPJchiMn7fL2eWIiIhcswIHrX379tG3b9886/v160dsbGyRFCUV15O9GgDwU8wR5m496uRqRERErk2Bg1Z4eDgLFizIs37+/PmEh4cXSVFScTWp7sfw9hEAjJmxkY0HTjm3IBERkWtQ4LsOn3jiCUaPHp3zIGmA5cuXM23aNN57770iL1Aqnv+7uREHTqaxcMcxHpi+jlmPdqBGFS9nlyUiIlJgBQ5ajzzyCMHBwfznP//hu+++A6BRo0bMmDGD/v37F3mBUvG42Kx8cFcLBn28kr+OJDF82hpmPtIefy83Z5cmIiJSIAW6dJiVlcXLL79M69atWbZsGSdOnODEiRMsW7ZMIQsYOHAglStXzneeMSmYSu4uTBnemlA/D/YdT+WhL9aTnqVnIYqISNlSoKDl4uLCW2+9RVZWVnHVU6Y99thjfP75584uo9wI8vVgyojW+Li7sDr2JE//sBnDMJxdloiIyFUr8GD47t27s2TJkuKopczr0qULPj4+zi6jXGkY7Muku1viYrUwO+YI72raBxERKUMKHLR69+7NM888w7hx4/jmm2/4+eefcy3X4o033sBisTBmzJhr2s7Fli5dSt++fQkNDcVisTB79ux8202cOJGIiAg8PDxo27Yta9asKdI6pHA61qvKawObAvD+wj18t+6gkysSERG5OgUeDP/oo48CMH78+DzvWSwW7PbCjaNZu3YtH3/8Mc2aNbtsu+XLl9OmTRtcXV1zrd+2bRtVqlQhKCgoz2dSU1OJiorivvvu49Zbb813uzNmzGDs2LFMnjyZtm3bMmHCBHr16sXOnTupVq0aAM2bN8/3sukff/xBaGjo1R6qFMKg1uEcPJXGBwv38NzMLYT6edKxXlVnlyUiInJZBe7Rcjgcl1wKG7JSUlIYOnQo//3vf6lcufJl9x0dHc2QIUNy7Wvnzp1069aN6dOn5/u53r1788orrzBw4MBLbnv8+PGMHDmSESNGEBkZyeTJk/Hy8mLKlCk5bWJiYti6dWueRSGrZIztWZ/+zUPJchg88uV6dh5NdnZJIiIil1Woh0oXtejoaG6++WZ69Ohx2XZWq5U5c+awceNG7r33XhwOB3v37qVbt24MGDCAp556qlD7z8jIYP369bn2b7Va6dGjBytXrizUNi9n4sSJREZG0rp16yLfdnlmsVh46/ZmtKkVQHJ6FiOmriE+6ayzyxIREbmkqw5aCxcuJDIykqSkpDzvJSYm0rhxY5YuXVrgAr799ls2bNjA66+/flXtQ0NDWbhwIcuWLWPIkCF069aNHj16MGnSpALvO1tCQgJ2uz3PZcegoCCOHr36x8D06NGDO+64gzlz5hAWFnbJkBYdHc22bdtYu3ZtoWuuqNxdbHxyT0tqB1biSOJZ7p++ltR03QUrIiKl01UHrQkTJjBy5Eh8fX3zvOfn58dDDz3Eu+++W6CdHzx4kMcee4yvvvoKDw+Pq/5cjRo1+OKLL5gxYwYuLi589tlnWCyWAu27OMyfP5/jx4+TlpbGoUOHaNeunbNLKpf8vdyYNrwNVSq5sfVwEmO+24xdsz6IiEgpdNVBa9OmTdx0002XfP/GG29k/fr1Bdr5+vXrOXbsGNdddx0uLi64uLiwZMkS3n//fVxcXC455is+Pp4HH3yQvn37kpaWxuOPP16g/V6satWq2Gw24uPj8+wnODj4mrYtxaNGFS/+O6wV7i5WFu9KYGasVXNsiYhIqXPVQSs+Pj7PnX4XcnFx4fjx4wXaeffu3dmyZQsxMTE5S6tWrRg6dCgxMTHYbLY8n0lISKB79+40atSImTNnsmDBAmbMmMG4ceMKtO8Lubm50bJly1wPy3Y4HCxYsEC9UqXYdTUqM2FwcywWWBZvZeqKv51dkoiISC5XPb1D9erV2bp1K3Xr1s33/c2bNxMSElKgnfv4+NCkSZNc6ypVqkSVKlXyrAcz/PTu3ZuaNWvmXDaMjIxk3rx5dOvWjerVq+fbu5WSksKePXtyXsfGxhITE0NAQAA1atQAYOzYsQwbNoxWrVrRpk0bJkyYQGpqKiNGjCjQMUnJ6t00hGd61ef1ubt44/dd1KjiTe+mBftzKCIiUlyuOmj16dOHf/3rX9x00015xlOdOXOGF154gVtuuaXIC7yQ1Wrltddeo1OnTri5nX/AcFRUFPPnzycwMDDfz61bt46uXbvmvB47diwAw4YNY9q0aQAMHjyY48eP8/zzz3P06FGaN2/O3Llz852XS0qXEe1rsmzjDv6MtzJmRgxBfh5cV+PS04SIiIiUlKsOWv/3f//HzJkzqV+/PqNGjaJBgwYA7Nixg4kTJ2K32/nnP/95zQUtXrz4su/37Nkz3/UtWrS45Ge6dOlyVeN3Ro0axahRo67YTkoXi8XCwFoOXPyrsWhnAiOnr2Pmo+2pWaWSs0sTEZEK7qrHaAUFBbFixQqaNGnCs88+y8CBAxk4cCDPPfccTZo0YdmyZer9EaexWeDdO5rRpLovJ1IzGDF1LafTMpxdloiIVHAFegRPzZo1mTNnDqdOnWLPnj0YhkG9evUuO5u7SEmp5O7ClGGtGTBxOfsSUnnw8/V88UAb3F3y3lQhIiJSEgo1M3zlypVp3bo1bdq0UciSUqWarwdTR7TBx92FNftP8uT3m3E4NO2DiIg4R6l4BI9IUWoQ7MOku1viYrXw86YjjJ+3y9kliYhIBaWgJeVSx3pVee3WpgB8uGgPM9YecHJFIiJSESloSbk1qFU4o7uZ8749N2srS3cVbEJdERGRa6WgJeXa4z3rM7BFdewOg0e/2sCOo3kfii4iIlJcFLSkXLNYLLxxW1Pa1gogJT2LEVPXEp901tlliYhIBaGgJeWeu4uNT+5pRZ3ASsQlnmXE1LWkpGc5uywREakAFLSkQvDzcmXaiDZU9XZjW1wSD36+jqSzmc4uS0REyjkFLakwwgO8+HRYazxdbazYe4JbP1rB3ydSnV2WiIiUYwpaUqE0D/fn+4fbEezrwZ5jKfSfuJxV+044uywRESmnFLSkwmlS3Y+fR3UgKsyP02mZ3PPZas2zJSIixUJBSyqkar4ezHioHbc0CyHTbvD0j1t49ddt2PW4HhERKUIKWlJhebja+OCuFozpUQ+A//4Zy8jP15GsQfIiIlJEFLSkQrNYLIzpUZ8P7mqBu4uVhTuOcdukFRw8mebs0kREpBxQ0BIB+kaF8t1D7ajm486ueHOQ/Nr9J51dloiIlHEKWiLnRIX789OoDjQO9eVkagZD/7uaH9YfcnZZIiJShiloiVwgxM+T7x9uR+8mwWTYHYz7fhNv/LYDhwbJi4hIIShoiVzEy82FiUOu4x/d6gIwecleHvpyPal6bI+IiBSQgpZIPqxWC0/c2ID37myOm4uVedviuX3ySg6fPuPs0kREpAxR0BK5jP7Nq/Ptg9dT1dud7XFJ9P9wORsOnHJ2WSIiUkYoaIlcwXU1KvPTqA40CvElISWdOz9ZxU8xh51dloiIlAEKWiJXobq/Jz883I4ejYLIyHLw2Lcx/OePnRokLyIil6WgJXKVKrm78Mk9LXm4cx0APli4h+ivN5CWoUHyIiKSPwUtkQKwWi0807sh79wRhavNwm9bjzLo45XEJWqQvIiI5KWgJVIIt7cM4+uR1xNQyY2th81B8psOnnZ2WSIiUsooaIkUUuuIAH6K7kD9IG+OJacz6OOV/LL5iLPLEhGRUkRBS+QahAd48eMj7enaIJD0LAejvt7IhPm7MAwNkhcREQUtkWvm4+HKp8Na80DHWgBMmL+bh79cT+KZTCdXJiIizqagJVIEbFYL/3dLJG/e1hRXm4Xf/4qn7wfL2Ho40dmliYiIEyloiRShwa1r8MPD7Qmr7MmBk2nc+tEKvlz1ty4liohUUApaIkUsKtyfX//RyZzc1O7g/2Zv5bFvY0jRQ6lFRCocBS2RYuDn5cp/723JP/s0wma18POmI/T7YBk7jiY5uzQRESlBCloixcRisTDyhtp899D1hPh5sC8hlQETl/PduoPOLk1EREqIgpZIMWtZM4BfR3eic/1AzmY6eOqHzYz7fhNnMuzOLk1ERIqZgpZICQio5MbU4a15slcDrBb4Yf0hBkxczp5jKc4uTUREipGClkgJsVotRHety1cPXE+gjzs745Pp9+Eyfoo57OzSRESkmChoFaGBAwdSuXJlbr/9dmeXIqVYuzpV+HV0R9rVrkJahp3Hvo3hn7O2cDZTlxJFRMobBa0i9Nhjj/H55587uwwpA6r5ePDlA20Z3a0uFgt8tfoAt01awd8nUp1dmoiIFCEFrSLUpUsXfHx8nF2GlBE2q4WxNzZg2og2BFRy468jSdzy/jLmbo1zdmkiIlJEnB60Jk2aRLNmzfD19cXX15d27drx22+/Fek+li5dSt++fQkNDcVisTB79ux8202cOJGIiAg8PDxo27Yta9asKdI6RPLTuX4gv47uSKualUlOz+LhLzfw0v/+IiPL4ezSRETkGjk9aIWFhfHGG2+wfv161q1bR7du3ejfvz9//fVXvu2XL19OZmbeh/Vu27aN+Pj4fD+TmppKVFQUEydOvGQdM2bMYOzYsbzwwgts2LCBqKgoevXqxbFjx3LaNG/enCZNmuRZjhw5UsCjFsktxM+Tbx68noc61wZg6vL93PHxSg6dSnNyZSIici2cHrT69u1Lnz59qFevHvXr1+fVV1/F29ubVatW5WnrcDiIjo5myJAh2O3nBw7v3LmTbt26MX369Hz30bt3b1555RUGDhx4yTrGjx/PyJEjGTFiBJGRkUyePBkvLy+mTJmS0yYmJoatW7fmWUJDQ6/hJyBicrVZebZ3Iz69txV+nq5sOniam99fxoLt+f8HQkRESj+nB60L2e12vv32W1JTU2nXrl2e961WK3PmzGHjxo3ce++9OBwO9u7dS7du3RgwYABPPfVUofabkZHB+vXr6dGjR6599ejRg5UrVxb6eC5l4sSJREZG0rp16yLftpR9PSKD+OUfHYkK8yPxTCb3T1/HG7/tIMuuS4kiImVNqQhaW7ZswdvbG3d3dx5++GFmzZpFZGRkvm1DQ0NZuHAhy5YtY8iQIXTr1o0ePXowadKkQu8/ISEBu91OUFBQrvVBQUEcPXr0qrfTo0cP7rjjDubMmUNYWNglQ1p0dDTbtm1j7dq1ha5ZyrfwAC++f7g9w9tHADB5yV6G/Hc1R5POOrcwEREpEBdnFwDQoEEDYmJiSExM5IcffmDYsGEsWbLkkmGrRo0afPHFF3Tu3JnatWvz2WefYbFYSrjqvObPn+/sEqQccXOx8mK/xrSpFcBTP2xmzf6T9P9oJYNrWOjj7OJEROSqlIoeLTc3N+rWrUvLli15/fXXiYqK4r333rtk+/j4eB588EH69u1LWloajz/++DXtv2rVqthstjyD6ePj4wkODr6mbYtcqz5NQ/jlHx2JDPHlZGomk7bbGPVNDNvjkpxdmoiIXEGpCFoXczgcpKen5/teQkIC3bt3p1GjRsycOZMFCxYwY8YMxo0bV+j9ubm50bJlSxYsWJCrhgULFuQ7VkykpEVUrcTMR9tzV+swAH7fdoze7/3Jg5+vY+vhRCdXJyIil+L0S4fPPvssvXv3pkaNGiQnJ/P111+zePFifv/99zxtHQ4HvXv3pmbNmsyYMQMXFxciIyOZN28e3bp1o3r16vn2bqWkpLBnz56c17GxscTExBAQEECNGjUAGDt2LMOGDaNVq1a0adOGCRMmkJqayogRI4rv4EUKwMPVxsv9IqmRsZ+t9jDm/HWUP7bF88e2eHo0qsZj3evTNMzP2WWKiMgFnB60jh07xr333ktcXBx+fn40a9aM33//nZ49e+Zpa7Vaee211+jUqRNubm4566Oiopg/fz6BgYH57mPdunV07do15/XYsWMBGDZsGNOmTQNg8ODBHD9+nOeff56jR4/SvHlz5s6dm2eAvIizhXrBA32a8fip+nywcA//23SE+duPMX/7Mbo1rMbo7vVoHu7v7DJFRIRSELQ+++yzArXPL4ABtGjR4pKf6dKlC4ZhXHHbo0aNYtSoUQWqR8RZ6lbz4b07WzC6ez0mLtzD7JjDLNxxjIU7jtG5fiCju9ejZc3Kzi5TRKRCK5VjtETk6tUJ9Gb84OYseKILt7cMw2a1sGTXcW6btIJ7PlvNuv0nnV2iiEiFpaAlUk7UqlqJd+6IYuETnRncKhwXq4U/dydw++SVDPnvKlbvO+HsEkVEKhwFLZFypmaVSrx5ezMWjevCXW1q4GqzsGLvCQZ/sorBH69kxd6Eq7qULiIi105BS6ScCg/w4vVbm7JoXBeGtjUD1+rYkwz572oGf7yKZbsVuEREipuClkg5F1bZi1cHNmXJk125t11N3GxW1uw/yd2freb2yStZuuu4ApeISDFR0BKpIEL9PXm5fxOWPtWV4e0jcHexsv7vU9w7ZQ0DP1rBop3HFLhERIqYgpZIBRPs58GL/Rrz51Ndub9jLTxcrcQcPM2IqWsZMHE5v22J42ym3dllioiUC06fR0tEnKOarwf/uiWShzvX4ZOle/ly1QE2HUrkka824O3uQo9G1bi5WSid6lXFw9Xm7HJFRMokBS2RCi7Qx51/3hzJQ53rMGVZLLM2HiYu8SyzY44wO+YIPu4u9IgM4uamIXSqXxV3F4UuEZGrpaAlIgBU9XbnqZsaMu7GBmw8eIpfNx9lzpY4jiadZdbGw8zaeBgfdxd6RgZxc7MQOtZT6BIRuRIFLRHJxWq10LJmAC1rBvB/Nzdiw4FT/Loljjlb4ohPSmfmxsPM3HgYHw8zdN3SLISOdQNxc9GQTxGRiyloicglWa0WWkUE0CoigH/dHMn6A6f4dbMZuo4lpzNzw2FmbjiMr4cLPSODuaVZCB3qVlXoEhE5R0FLRK6K1WqhdUQArSMCeP6WSNb9fYo5W86Hrh83HOLHDYfw9XDhxsbB3NwshA51FLpEpGJT0BKRArNaLbSpFUCbWgH865ZI1u0/aYaurUc5npzOD+sP8cP6Q/h5unLjuTFdHepWxdWm0CUiFYuClohcE5vVQtvaVWhbuwrP923M2uzQteUoCSnpfL/+EN+vP4S/lxm6bm8ZTuuIylgsFmeXLiJS7BS0RKTI2KwWrq9dhetrV+GFvo1ZE2uGrt+2xpGQksF36w7x3bpDNK3ux/0da3FzsxD1colIuaagJSLFwma10K5OFdrVqcKL/RqzOvYEP208wuyYw2w5nMiYGTG88dsO7m1fkyFtauDv5ebskkVEipz+Kykixc5mtdC+TlXevL0ZK57pxhM961PV252jSWd5a+5O2r2+kH/N3sq+4ynOLlVEpEgpaIlIiari7c4/utdj+TNdeeeOKBqF+HIm084Xq/6m+/glPDB9LSv2JugB1yJSLujSoYg4hbuLjdtbhnHbddVZufcEny2LZcGOY8zfbi6RIb4Mb1cDm8PZlYqIFJ6Clog4lcVioX3dqrSvW5W9x1OYujyWH9YfYltcEk/N3Iqvq41D3vu4p30tAippHJeIlC26dCgipUadQG9eGdCUlc9058leDQjycScp08K7C/bQ7vUFPDdrC3uOaRyXiJQdCloiUupUruRGdNe6LBzbiXvq2mkc6kN6loOvVx+gx/glDJ+6hmW7NY5LREo/XToUkVLLzcVKq0CDf/W+no2Hkvl0WSzzt8ezeOdxFu88TsNgH+7rWIt+UaF4uNqcXa6ISB4KWiJS6lks52ef35+QyrQV+/lu3UF2HE3mqR8289bcHdxzfQR3X1+DKt7uzi5XRCSHgpaIlCkRVSvxYr/GPN6jPt+uPcC0FfuJSzzLu/N3MXHRHpqG+REV5k9UuPm1ZhUvPe5HRJxGQUtEyiQ/L1ce6lyH+zrW4retR/nsz31sOpTI+r9Psf7vU+fbebrS7Fz4ahbmR1S4P0G+Hk6sXEQqEgUtESnTXG1W+kWF0rdZCLEJqWw6dJpNBxPZdOg0fx1JIvFMJn/uTuDP3Qk5nwn29cgJXVFh/jQN88PP09WJRyEi5ZWCloiUCxaLhdqB3tQO9GZgizAAMrIc7IpPZtOh02w+F752xSdzNOksR7ed5Y9t8Tmfr1W1ElFhfjQ7d9mxcaifBtiLyDVT0BKRcsvNxUqT6n40qe7H0LbmurSMLP46ksSmg6fZdCiRTQdPc+BkGrEJqcQmpDI75ggALlYL9YN8csZ6NQvzp36QNy42zYojIldPQUtEKhQvNxdaRwTQOiIgZ92p1Aw2HzZD1+ZDp4k5mEhCSjrb4pLYFpfEN2sOnvusjetrV6FTvap0qhdIncBKGmgvIpeloCUiFV7lSm50rh9I5/qBABiGQVziWTYfOt/rteVQIsnpWSzccYyFO44BEOLnQad6VelYL5COdavqEUEikoeClojIRSwWC6H+noT6e3JTkxAAHA6DHUeT+XP3cf7cncCa/SeJSzzLd+sO8d26Q1gs0DjUl071AulUtyotIyrj7qIxXiIVnYKWiMhVsFotRIb6Ehnqy0Od63A2086a2JMs25PA0l3H2XE0ma2Hk9h6OIlJi/fi6WqjTa2AnMuM9YO8dZlRpAJS0BIRKQQPVxs31A/khvqBPNenEceSz7J8TwJ/7krgzz0JHE9OZ8mu4yzZdRzYTjUfdzrWq8oN9QLpULcqgT6awV6kIlDQEhEpAtV8PBjYIoyBLcIwDIOd8cks253A0t0JrN53gmPJ6czccJiZGw4D0CjElxvqVaVjvaq0jghAFxlFyicFLRGRImaxWGgY7EvDYF8e6FSbs5l21v99iqW7j7NsdwJ/HUlie5y5fLx0H+4uVlrVrIzXWSsJqw4Q7OdFoI97zlLJzabLjiJllIKWiEgx83C10aFuVTrUrQq9ISEl3bzMuDuBZbsTOJp0luV7TwBW5h3ekc/nrWbo8jaDV1Xv8yHs4nWaZFWkdFHQEhEpYVW93enfvDr9m1fHMAz2HEth8c54lq7fTqWqIZxIzeB4cjrHk9NJzbBzNtPBwZNnOHjyzBW37ePhkjuMXRDKmof7U6+aBuWLlCQFLRERJ7JYLNQL8iEiwINqp/6iT58oXF3PP3cxLSOLhOQMjqeczQlfx5PTOZ5yLoylpJNw7mtGloPks1kkn81i3/HUfPcXVtmTrg2q0a1hNdrVqaIeMJFipqAlIlKKebm5UKOKCzWqeF22nWEYJJ3NuiCInQ9gx5PTOXzqDOsPnOLQqTN8sepvvlj1Nx6uVtrXqUrXBoF0bViNsMqX34eIFJyClohIOWCxWPDzdMXP05W61bzzbZOWkcWKPSdYuPMYi3YcIy7x7PmZ7n/6i/pB3nRtWI1uDarRsmZlPddRpAgoaImIVBBebi70iAyiR2RQzhQUC3eYoWv936fYFZ/CrvgUPl6yD18PF26oH0jXBtXo0iCQKt6a90ukMBS0REQqoAunoHi0S11Op2WwZNdxFu88zuKdxziVlskvm+P4ZXMcFgtEhfnTraE5tqtxqK8G1ItcJQUtERHB38st505Iu8Mg5uBpFp27rLgtLomYg6eJOXia8fN2Uc3HnS4NAunWsBod6wXi7a5/SkQuRb8dIiKSi81qoWXNyrSsWZlxvRpwNPEsi3eaoWvZngSOJafnPEzb1WahTa0AujaoRsd6VWkQ5KPeLpELKGiJiMhlBft5cGebGtzZpgbpWebDtLPHdu0/kcbyPSdYvucEAFW93WhXpyod6lShQ92qhAfoTkap2BS0RETkqrm72OhUL5BO9QJ5oW9jYhNSWbjjGEt2HWdt7EkSUjL436Yj/G/TEQDCAzzpUMecFb99nSoaVC8VjoKWiIgUWq2qlbi/Yy3u71iL9Cw7Gw+cZsWeBJbvPcGmg6c5ePIM3548yLdrDwLQMNiHdrUDcD1l4Yb0LCpfMDmrSHmkoCUiIkXC3cXG9bWrcH3tKowFUtKzWBN74tylxQR2HE3OWcDGlNcWERXuT4c6VWhftyotavjj7qKZ6qV8UdASEZFi4e3uQreGQXRrGASYD9NeufcEy3YfY/6WQ5xIh/V/n2L936d4f+EePFyttI4IMB/AXacqkaG+2KwaWC9lm4KWiIiUiKre7vSNCuWmyEDau/5Ns3ZdWfP3aZbvOcGKvQkkpGTw5+4E/tydAICfpyvtalehQ90qNA+vjI+HC17uNiq5ueDlZtPdjVImKGiJiIhThFX2pFY1Xwa3roFhGOyKT2H5ngRW7E1g1b6TJJ7JZO5fR5n719E8n7VYwMvVhpe7C97uZvCq5HYuiLm7UMnNhpebC5VyXpttvN1d8LrgffO1jcpebuo9k2KhoCUiIk5nsVhoEOxDg2Af7utYiyy7g82HE82B9XtOsOd4CmnpWaRm2AEwDEjNsJOaYed4cvo179/Hw4W2tQJyxpg1CtFlSykaCloiIlLquNisXFejMtfVqMyobvVy1jscBmez7KSm20lNzyI1I4u0jHPfp9vN1+cCWVpGVk67tAzzvex2aRnn2pxrm3w2i/nbjzF/+zFAwUuKjoKWiIiUGVarBS83F7zcXAj0KZo5ubLsDrbHJbNyn3nJcm3sSQUvKTIKWiIiUqG52Kw0DfOjaZgfD95Qhyy7g21xSazad4JV+06yJp/g5evhQptaVbi+doCCl1yWgpaIiMgFXGxWmoX50yzMP1fwWrn3BKv2nWDt/lMknc1i/vZ45m+PBxS85NIUtERERC7jwuD1UGczeP11JLvH68rBq1UNP86N4ZcKSEGrCA0cOJDFixfTvXt3fvjhB2eXIyIixcDFZiUq3J+o8KsPXuDCa1sXUb2yJyF+nlT39yTU34MQP09Cz31fzcdDvWDlkIJWEXrssce47777mD59urNLERGREpJf8Np6QfBat/8UKelZnErL5FRaJlsPJ+W/HauFIF8PQv09CPXPDmTnw1h1f098PV00UWsZo6BVhLp06cLixYudXYaIiDiRi81K83B/mof783DnOmRkZPDj/36jcetOHE/N5MjpMxw+fZa4xDMcOX2GI6fPcjTpLFkOg8Onz3D49BngVL7b9nKznQthHlQ/F8aC/dzx8XDNmZC10rlJXCudm8jV3cWqcOZETg9ar7/+OjNnzmTHjh14enrSvn173nzzTRo0aFBk+1i6dClvv/0269evJy4ujlmzZjFgwIA87SZOnMjbb7/N0aNHiYqK4oMPPqBNmzZFVoeIiFQ8FosFLxdoFOJDM1fXfNvYHQbHk9M5fNoMX2YIO2sGsXPfn0zNIC3Dzp5jKew5lnLV+3exWi4IX+dnxL9w1nzz/XOvc9aZwa2ajwfhAZ4Ka4Xk9KC1ZMkSoqOjad26NVlZWTz33HPceOONbNu2jUqVKuVpv3z5ctq0aYPrRX9Yt23bRpUqVQgKCsrzmdTUVKKiorjvvvu49dZb861jxowZjB07lsmTJ9O2bVsmTJhAr1692LlzJ9WqVQOgefPmZGVl5fnsH3/8QWhoaGEOX0REBJvVQrCfB8F+HrSsWTnfNmcy7MQlniEu8SyHT58h7lwQi08+m2vC1tT0LFLSszib6QAgy2GQeCaTxDOZha7Px8OFJqHmFBhNqvvRJNSXiCqVsGpM2RU5PWjNnTs31+tp06ZRrVo11q9fzw033JDrPYfDQXR0NPXq1ePbb7/FZrMBsHPnTrp168bYsWN56qmn8uyjd+/e9O7d+7J1jB8/npEjRzJixAgAJk+ezK+//sqUKVN45plnAIiJiSnsYYqIiFwTTzcbtQO9qR3ofVXts+wO0jKzZ83PIiX9/PepGedfp13wXsq5oJaWbiflXLu402dJPpvFyn0nWLnvRM72fdxdiAz1pUl1P5pWNwNYuJ9bcR1+meX0oHWxxMREAAICAvK8Z7VamTNnDjfccAP33nsvX3zxBbGxsXTr1o0BAwbkG7KuRkZGBuvXr+fZZ5/Nta8ePXqwcuXKwh3IZUycOJGJEydit+t+XxERKR4uNiu+Niu+HvlfrrxamXYHu+KT2Xo4ka2Hk9hyOJHtcUkkp2exOvYkq2NP5rSt5GYjyN3GBnYQFV6ZptX9qB3oXaHvpixVQcvhcDBmzBg6dOhAkyZN8m0TGhrKwoUL6dSpE0OGDGHlypX06NGDSZMmFXq/CQkJ2O32PJcdg4KC2LFjx1Vvp0ePHmzatInU1FTCwsL4/vvvadeuXZ520dHRREdHk5SUhJ+fX6HrFhERKW6uNiuNQ/1oHOrH4Nbmuky7gz3HUs6Fr0S2HE5kW1wSqRl29mVY2LfyAKw8AICnq43IUN+cXq8m1X2pG+iNi83qxKMqOaUqaEVHR7N161aWLVt22XY1atTgiy++oHPnztSuXZvPPvusVAzSmz9/vrNLEBERKXauNiuNQnxpFOLLHa3CAfNS5a64RL6a+ye2qrXYfjSZv44kkZZhZ/3fp1j/9/k7KT1czc83CfWjmo87NpsFF6sFm9V67qvl/Fdb/utdbdbc7azWC9qfX+/j7oqf17X16l2LUhO0Ro0axS+//MLSpUsJCwu7bNv4+HgefPBB+vbty9q1a3n88cf54IMPCr3vqlWrYrPZiI+Pz7U+Pj6e4ODgQm9XRESkonCxWakX5E2bQIM+fRri6uqK3WEQm5DClgsuO247kkRKehYbD5xm44HTxV7XXW1q8PqtTYt9P5fi9KBlGAb/+Mc/mDVrFosXL6ZWrVqXbZ+QkED37t1p1KgR33//Pbt27aJLly64u7vzzjvvFKoGNzc3WrZsyYIFC3KmfXA4HCxYsIBRo0YVapsiIiIVnc1qoW41H+pW82FgC3Odw2Gw/0RqTuhKOpuF3eEgy2FgdxhkOQyy7I6c7+0Ogyx79nsXrXdcsN6e/3p3F+deonR60IqOjubrr7/mp59+wsfHh6NHjwLg5+eHp6dnrrYOh4PevXtTs2ZNZsyYgYuLC5GRkcybN49u3bpRvXp1Hn/88Tz7SElJYc+ePTmvY2NjiYmJISAggBo1agAwduxYhg0bRqtWrWjTpg0TJkwgNTU15y5EERERuXZWqyXn7sn+zas7u5xi5/SglT2IvUuXLrnWT506leHDh+daZ7Vaee211+jUqRNubudvIY2KimL+/PkEBgbmu49169bRtWvXnNdjx44FYNiwYUybNg2AwYMHc/z4cZ5//nmOHj1K8+bNmTt3br7zcomIiIhcDacHLcMwCtS+Z8+e+a5v0aLFJT/TpUuXq9rPqFGjdKlQREREikzFuLdSRERExAkUtERERESKiYKWiIiISDFR0BIREREpJgpaIiIiIsVEQUtERESkmChoiYiIiBQTBS0RERGRYqKgJSIiIlJMFLREREREiomCloiIiEgxUdASERERKSZOf6h0RZb9oOukpKQi33ZmZiZpaWkkJSXh6upa5NsvTXSs5VdFOl4da/lVkY63ohxr9r/b2f+OX46ClhMlJycDEB4e7uRKREREpKCSk5Px8/O7bBuLcTVxTIqFw+HgyJEj+Pj4YLFYinTbSUlJhIeHc/DgQXx9fYt026WNjrX8qkjHq2MtvyrS8VaUYzUMg+TkZEJDQ7FaLz8KSz1aTmS1WgkLCyvWffj6+pbrP+wX0rGWXxXpeHWs5VdFOt6KcKxX6snKpsHwIiIiIsVEQUtERESkmCholVPu7u688MILuLu7O7uUYqdjLb8q0vHqWMuvinS8FelYr5YGw4uIiIgUE/VoiYiIiBQTBS0RERGRYqKgJSIiIlJMFLREREREiomCVhk2ceJEIiIi8PDwoG3btqxZs+ay7b///nsaNmyIh4cHTZs2Zc6cOSVUaeG9/vrrtG7dGh8fH6pVq8aAAQPYuXPnZT8zbdo0LBZLrsXDw6OEKi68F198MU/dDRs2vOxnyuI5zRYREZHneC0WC9HR0fm2L0vndenSpfTt25fQ0FAsFguzZ8/O9b5hGDz//POEhITg6elJjx492L179xW3W9Df+ZJwuWPNzMzk6aefpmnTplSqVInQ0FDuvfdejhw5ctltFuZ3oaRc6dwOHz48T+033XTTFbdb1s4tkO/vr8Vi4e23377kNkvzuS0uClpl1IwZMxg7diwvvPACGzZsICoqil69enHs2LF8269YsYK77rqL+++/n40bNzJgwAAGDBjA1q1bS7jyglmyZAnR0dGsWrWKefPmkZmZyY033khqauplP+fr60tcXFzO8vfff5dQxdemcePGuepetmzZJduW1XOabe3atbmOdd68eQDccccdl/xMWTmvqampREVFMXHixHzff+utt3j//feZPHkyq1evplKlSvTq1YuzZ89ecpsF/Z0vKZc71rS0NDZs2MC//vUvNmzYwMyZM9m5cyf9+vW74nYL8rtQkq50bgFuuummXLV/8803l91mWTy3QK5jjIuLY8qUKVgsFm677bbLbre0nttiY0iZ1KZNGyM6Ojrntd1uN0JDQ43XX3893/aDBg0ybr755lzr2rZtazz00EPFWmdRO3bsmAEYS5YsuWSbqVOnGn5+fiVXVBF54YUXjKioqKtuX17OabbHHnvMqFOnjuFwOPJ9v6yeV8CYNWtWzmuHw2EEBwcbb7/9ds6606dPG+7u7sY333xzye0U9HfeGS4+1vysWbPGAIy///77km0K+rvgLPkd77Bhw4z+/fsXaDvl5dz279/f6Nat22XblJVzW5TUo1UGZWRksH79enr06JGzzmq10qNHD1auXJnvZ1auXJmrPUCvXr0u2b60SkxMBCAgIOCy7VJSUqhZsybh4eH079+fv/76qyTKu2a7d+8mNDSU2rVrM3ToUA4cOHDJtuXlnIL5Z/rLL7/kvvvuu+wD1svqeb1QbGwsR48ezXXu/Pz8aNu27SXPXWF+50urxMRELBYL/v7+l21XkN+F0mbx4sVUq1aNBg0a8Mgjj3DixIlLti0v5zY+Pp5ff/2V+++//4pty/K5LQwFrTIoISEBu91OUFBQrvVBQUEcPXo0388cPXq0QO1LI4fDwZgxY+jQoQNNmjS5ZLsGDRowZcoUfvrpJ7788kscDgft27fn0KFDJVhtwbVt25Zp06Yxd+5cJk2aRGxsLJ06dSI5OTnf9uXhnGabPXs2p0+fZvjw4ZdsU1bP68Wyz09Bzl1hfudLo7Nnz/L0009z1113XfaBwwX9XShNbrrpJj7//HMWLFjAm2++yZIlS+jduzd2uz3f9uXl3E6fPh0fHx9uvfXWy7Yry+e2sFycXYDI1YqOjmbr1q1XvJ7frl072rVrl/O6ffv2NGrUiI8//ph///vfxV1mofXu3Tvn+2bNmtG2bVtq1qzJd999d1X/SyzLPvvsM3r37k1oaOgl25TV8yqmzMxMBg0ahGEYTJo06bJty/Lvwp133pnzfdOmTWnWrBl16tRh8eLFdO/e3YmVFa8pU6YwdOjQK96gUpbPbWGpR6sMqlq1Kjabjfj4+Fzr4+PjCQ4OzvczwcHBBWpf2owaNYpffvmFRYsWERYWVqDPurq60qJFC/bs2VNM1RUPf39/6tevf8m6y/o5zfb3338zf/58HnjggQJ9rqye1+zzU5BzV5jf+dIkO2T9/fffzJs377K9Wfm50u9CaVa7dm2qVq16ydrL+rkF+PPPP9m5c2eBf4ehbJ/bq6WgVQa5ubnRsmVLFixYkLPO4XCwYMGCXP/jv1C7du1ytQeYN2/eJduXFoZhMGrUKGbNmsXChQupVatWgbdht9vZsmULISEhxVBh8UlJSWHv3r2XrLusntOLTZ06lWrVqnHzzTcX6HNl9bzWqlWL4ODgXOcuKSmJ1atXX/LcFeZ3vrTIDlm7d+9m/vz5VKlSpcDbuNLvQml26NAhTpw4ccnay/K5zfbZZ5/RsmVLoqKiCvzZsnxur5qzR+NL4Xz77beGu7u7MW3aNGPbtm3Ggw8+aPj7+xtHjx41DMMw7rnnHuOZZ57Jab98+XLDxcXFeOedd4zt27cbL7zwguHq6mps2bLFWYdwVR555BHDz8/PWLx4sREXF5ezpKWl5bS5+Fhfeukl4/fffzf27t1rrF+/3rjzzjsNDw8P46+//nLGIVy1J554wli8eLERGxtrLF++3OjRo4dRtWpV49ixY4ZhlJ9zeiG73W7UqFHDePrpp/O8V5bPa3JysrFx40Zj48aNBmCMHz/e2LhxY86ddm+88Ybh7+9v/PTTT8bmzZuN/v37G7Vq1TLOnDmTs41u3boZH3zwQc7rK/3OO8vljjUjI8Po16+fERYWZsTExOT6HU5PT8/ZxsXHeqXfBWe63PEmJycb48aNM1auXGnExsYa8+fPN6677jqjXr16xtmzZ3O2UR7ObbbExETDy8vLmDRpUr7bKEvntrgoaJVhH3zwgVGjRg3Dzc3NaNOmjbFq1aqc9zp37mwMGzYsV/vvvvvOqF+/vuHm5mY0btzY+PXXX0u44oID8l2mTp2a0+biYx0zZkzOzyUoKMjo06ePsWHDhpIvvoAGDx5shISEGG5ubkb16tWNwYMHG3v27Ml5v7yc0wv9/vvvBmDs3Lkzz3tl+bwuWrQo3z+32cfjcDiMf/3rX0ZQUJDh7u5udO/ePc/PoGbNmsYLL7yQa93lfued5XLHGhsbe8nf4UWLFuVs4+JjvdLvgjNd7njT0tKMG2+80QgMDDRcXV2NmjVrGiNHjswTmMrDuc328ccfG56ensbp06fz3UZZOrfFxWIYhlGsXWYiIiIiFZTGaImIiIgUEwUtERERkWKioCUiIiJSTBS0RERERIqJgpaIiIhIMVHQEhERESkmCloiIiIixURBS0TEiSIiIpgwYYKzyxCRYqKgJSIVxvDhwxkwYAAAXbp0YcyYMSW272nTpuHv759n/dq1a3nwwQdLrA4RKVkuzi5ARKQsy8jIwM3NrdCfDwwMLMJqRKS0UY+WiFQ4w4cPZ8mSJbz33ntYLBYsFgv79+8HYOvWrfTu3Rtvb2+CgoK45557SEhIyPlsly5dGDVqFGPGjKFq1ar06tULgPHjx9O0aVMqVapEeHg4jz76KCkpKQAsXryYESNGkJiYmLO/F198Ech76fDAgQP0798fb29vfH19GTRoEPHx8Tnvv/jiizRv3pwvvviCiIgI/Pz8uPPOO0lOTi7eH5qIFIqClohUOO+99x7t2rVj5MiRxMXFERcXR3h4OKdPn6Zbt260aNGCdevWMXfuXOLj4xk0aFCuz0+fPh03NzeWL1/O5MmTAbBarbz//vv89ddfTJ8+nYULF/LUU08B0L59eyZMmICvr2/O/saNG5enLofDQf/+/Tl58iRLlixh3rx57Nu3j8GDB+dqt3fvXmbPns0vv/zCL7/8wpIlS3jjjTeK6aclItdClw5FpMLx8/PDzc0NLy8vgoODc9Z/+OGHtGjRgtdeey1n3ZQpUwgPD2fXrl3Ur18fgHr16vHWW2/l2uaF470iIiJ45ZVXePjhh/noo49wc3PDz88Pi8WSa38XW7BgAVu2bCE2Npbw8HAAPv/8cxo3bszatWtp3bo1YAayadOm4ePjA8A999zDggULePXVV6/tByMiRU49WiIi52zatIlFixbh7e2dszRs2BAwe5GytWzZMs9n58+fT/fu3alevTo+Pj7cc889nDhxgrS0tKve//bt2wkPD88JWQCRkZH4+/uzffv2nHURERE5IQsgJCSEY8eOFehYRaRkqEdLROSclJQU+vbty5tvvpnnvZCQkJzvK1WqlOu9/fv3c8stt/DII4/w6quvEhAQwLJly7j//vvJyMjAy8urSOt0dXXN9dpiseBwOIp0HyJSNBS0RKRCcnNzw26351p33XXX8eOPPxIREYGLy9X/9bh+/XocDgf/+c9/sFrNCwXffffdFfd3sUaNGnHw4EEOHjyY06u1bds2Tp8+TWRk5FXXIyKlhy4dikiFFBERwerVq9m/fz8JCQk4HA6io6M5efIkd911F2vXrmXv3r38/vvvjBgx4rIhqW7dumRmZvLBBx+wb98+vvjii5xB8hfuLyUlhQULFpCQkJDvJcUePXrQtGlThg4dyoYNG1izZg333nsvnTt3plWrVkX+MxCR4qegJSIV0rhx47DZbERGRhIYGMiBAwcIDQ1l+fLl2O12brzxRpo2bcqYMWPw9/fP6anKT1RUFOPHj+fNN9+kSZMmfPXVV7z++uu52rRv356HH36YwYMHExgYmGcwPZiXAH/66ScqV67MDTfcQI8ePahduzYzZswo8uMXkZJhMQzDcHYRIiIiIuWRerREREREiomCloiIiEgxUdASERERKSYKWiIiIiLFREFLREREpJgoaImIiIgUEwUtERERkWKioCUiIiJSTBS0RERERIqJgpaIiIhIMVHQEhERESkmCloiIiIixeT/AexHr436MnAfAAAAAElFTkSuQmCC",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# vizualize mnist\n",
    "\n",
    "for item in history.history:\n",
    "    print(\"Key:\",item)\n",
    "\n",
    "plt.plot(history.history['loss'],label=\"train\")\n",
    "plt.plot(history.history['val_loss'],label=\"validation\")\n",
    "plt.title('Model Loss')\n",
    "plt.yscale('log')\n",
    "plt.ylabel('Cross Entropy')\n",
    "plt.xlabel('Iteration')\n",
    "plt.grid()\n",
    "plt.legend()\n",
    "plt.show()\n",
    "\n",
    "#print(\"history\",history.history)\n",
    "plt.plot(history.history['sparse_categorical_accuracy'],label=\"train\")\n",
    "plt.plot(history.history['val_sparse_categorical_accuracy'],label=\"validation\")\n",
    "plt.title('Model Accuracy')\n",
    "#plt.yscale('log')\n",
    "plt.ylabel('Acc')\n",
    "plt.xlabel('Iteration')\n",
    "plt.grid()\n",
    "plt.legend()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 5.\n",
    "\n",
    "Esant persimokymui papildykite modelius Dropout sluoksniais.\n",
    "\n",
    "Ats.:\n",
    "Neatrodo, kad yra persimokymas. Atrodo, kad kažkas atvirkčia persimokymo vyksta, stipriai pakėlus \"Dropout\" validacijos tikslumas yra ženkliai aukštesnis negu mokymosi. Kodėl?\n",
    "\n",
    "Kas dar vyksta didinant \"Dropout\" tai maksimalus mokymosi tikslumas žemėja šiek tiek."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 6.\n",
    "\n",
    "Palyginkite Lab24 MNIST ir architektūros su konvoliuciniais sluoksniais tikslumą bei mokymosi greitį MNIST duomenims.\n",
    "\n",
    "Ats.: Mokymosi laikas ženkliai pailgėja, bet dėsningumus ir sąryšius kuriuos mokymose proceses gali išmokti atstoją ilgesnį mokymasi.\n",
    "\n",
    "Jeigu reikia greičio gali naudoti \"Batching\", kad pagreitinti mokymasi."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Lab32"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 1.\n",
    "\n",
    "Realizuoti LeNet architektūrą ir pritaikyti MNIST duomenims."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 58,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model: \"sequential_36\"\n",
      "_________________________________________________________________\n",
      " Layer (type)                Output Shape              Param #   \n",
      "=================================================================\n",
      " zero_padding2d_8 (ZeroPadd  (None, 32, 32, 1)         0         \n",
      " ing2D)                                                          \n",
      "                                                                 \n",
      " conv2d_75 (Conv2D)          (None, 28, 28, 6)         156       \n",
      "                                                                 \n",
      " average_pooling2d_16 (Aver  (None, 14, 14, 6)         0         \n",
      " agePooling2D)                                                   \n",
      "                                                                 \n",
      " conv2d_76 (Conv2D)          (None, 10, 10, 16)        2416      \n",
      "                                                                 \n",
      " average_pooling2d_17 (Aver  (None, 5, 5, 16)          0         \n",
      " agePooling2D)                                                   \n",
      "                                                                 \n",
      " flatten_35 (Flatten)        (None, 400)               0         \n",
      "                                                                 \n",
      " dense_76 (Dense)            (None, 120)               48120     \n",
      "                                                                 \n",
      " dense_77 (Dense)            (None, 84)                10164     \n",
      "                                                                 \n",
      " dense_78 (Dense)            (None, 10)                850       \n",
      "                                                                 \n",
      "=================================================================\n",
      "Total params: 61706 (241.04 KB)\n",
      "Trainable params: 61706 (241.04 KB)\n",
      "Non-trainable params: 0 (0.00 Byte)\n",
      "_________________________________________________________________\n",
      "Epoch 1/20\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "f:\\KTU\\Neuroninių tinklų metodai\\venv\\Lib\\site-packages\\keras\\src\\backend.py:5727: UserWarning: \"`sparse_categorical_crossentropy` received `from_logits=True`, but the `output` argument was produced by a Softmax activation and thus does not represent logits. Was this intended?\n",
      "  output, from_logits = _get_logits(\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "600/600 [==============================] - 3s 4ms/step - loss: 2.3026 - sparse_categorical_accuracy: 0.1172 - val_loss: 2.2874 - val_sparse_categorical_accuracy: 0.1135\n",
      "Epoch 2/20\n",
      "600/600 [==============================] - 2s 4ms/step - loss: 2.1563 - sparse_categorical_accuracy: 0.3599 - val_loss: 1.7558 - val_sparse_categorical_accuracy: 0.6306\n",
      "Epoch 3/20\n",
      "600/600 [==============================] - 2s 3ms/step - loss: 1.2294 - sparse_categorical_accuracy: 0.7310 - val_loss: 0.8575 - val_sparse_categorical_accuracy: 0.8134\n",
      "Epoch 4/20\n",
      "600/600 [==============================] - 2s 3ms/step - loss: 0.7044 - sparse_categorical_accuracy: 0.8393 - val_loss: 0.5652 - val_sparse_categorical_accuracy: 0.8741\n",
      "Epoch 5/20\n",
      "600/600 [==============================] - 2s 4ms/step - loss: 0.5064 - sparse_categorical_accuracy: 0.8783 - val_loss: 0.4372 - val_sparse_categorical_accuracy: 0.8958\n",
      "Epoch 6/20\n",
      "600/600 [==============================] - 3s 4ms/step - loss: 0.4106 - sparse_categorical_accuracy: 0.8958 - val_loss: 0.3699 - val_sparse_categorical_accuracy: 0.9074\n",
      "Epoch 7/20\n",
      "600/600 [==============================] - 2s 4ms/step - loss: 0.3561 - sparse_categorical_accuracy: 0.9062 - val_loss: 0.3270 - val_sparse_categorical_accuracy: 0.9136\n",
      "Epoch 8/20\n",
      "600/600 [==============================] - 2s 4ms/step - loss: 0.3199 - sparse_categorical_accuracy: 0.9138 - val_loss: 0.2981 - val_sparse_categorical_accuracy: 0.9195\n",
      "Epoch 9/20\n",
      "600/600 [==============================] - 3s 4ms/step - loss: 0.2928 - sparse_categorical_accuracy: 0.9196 - val_loss: 0.2759 - val_sparse_categorical_accuracy: 0.9266\n",
      "Epoch 10/20\n",
      "600/600 [==============================] - 2s 4ms/step - loss: 0.2721 - sparse_categorical_accuracy: 0.9248 - val_loss: 0.2574 - val_sparse_categorical_accuracy: 0.9306\n",
      "Epoch 11/20\n",
      "600/600 [==============================] - 2s 4ms/step - loss: 0.2553 - sparse_categorical_accuracy: 0.9294 - val_loss: 0.2401 - val_sparse_categorical_accuracy: 0.9341\n",
      "Epoch 12/20\n",
      "600/600 [==============================] - 2s 4ms/step - loss: 0.2411 - sparse_categorical_accuracy: 0.9320 - val_loss: 0.2274 - val_sparse_categorical_accuracy: 0.9382\n",
      "Epoch 13/20\n",
      "600/600 [==============================] - 2s 4ms/step - loss: 0.2287 - sparse_categorical_accuracy: 0.9370 - val_loss: 0.2179 - val_sparse_categorical_accuracy: 0.9387\n",
      "Epoch 14/20\n",
      "600/600 [==============================] - 2s 4ms/step - loss: 0.2178 - sparse_categorical_accuracy: 0.9388 - val_loss: 0.2090 - val_sparse_categorical_accuracy: 0.9421\n",
      "Epoch 15/20\n",
      "600/600 [==============================] - 2s 4ms/step - loss: 0.2082 - sparse_categorical_accuracy: 0.9408 - val_loss: 0.1997 - val_sparse_categorical_accuracy: 0.9441\n",
      "Epoch 16/20\n",
      "600/600 [==============================] - 2s 4ms/step - loss: 0.1997 - sparse_categorical_accuracy: 0.9435 - val_loss: 0.1935 - val_sparse_categorical_accuracy: 0.9463\n",
      "Epoch 17/20\n",
      "600/600 [==============================] - 2s 4ms/step - loss: 0.1918 - sparse_categorical_accuracy: 0.9452 - val_loss: 0.1856 - val_sparse_categorical_accuracy: 0.9492\n",
      "Epoch 18/20\n",
      "600/600 [==============================] - 2s 4ms/step - loss: 0.1846 - sparse_categorical_accuracy: 0.9479 - val_loss: 0.1802 - val_sparse_categorical_accuracy: 0.9498\n",
      "Epoch 19/20\n",
      "600/600 [==============================] - 2s 4ms/step - loss: 0.1783 - sparse_categorical_accuracy: 0.9494 - val_loss: 0.1744 - val_sparse_categorical_accuracy: 0.9508\n",
      "Epoch 20/20\n",
      "600/600 [==============================] - 2s 4ms/step - loss: 0.1725 - sparse_categorical_accuracy: 0.9507 - val_loss: 0.1691 - val_sparse_categorical_accuracy: 0.9527\n"
     ]
    }
   ],
   "source": [
    "mnist = tf.keras.datasets.mnist\n",
    "(train_images0, train_labels0), (test_images0, test_labels0) = mnist.load_data()\n",
    "\n",
    "test_images=test_images0\n",
    "train_images=train_images0\n",
    "\n",
    "test_images  = test_images/255.0\n",
    "train_images = train_images/255.0\n",
    "\n",
    "keras_model = tf.keras.models.Sequential([\n",
    "    layers.ZeroPadding2D(padding=(2,2), input_shape=(28, 28, 1)),\n",
    "    layers.Conv2D(6, (5, 5), activation='relu'),\n",
    "    layers.AveragePooling2D((2, 2), strides = 2),\n",
    "    layers.Conv2D(16, (5, 5)),\n",
    "    layers.AveragePooling2D((2, 2), strides = 2),\n",
    "\n",
    "    layers.Flatten(),\n",
    "    layers.Dense(120, activation='sigmoid'),\n",
    "    layers.Dense(84, activation='sigmoid'),\n",
    "    layers.Dense(10, activation='sigmoid')\n",
    "])\n",
    "\n",
    "keras_model.build(input_shape=[None,784])\n",
    "\n",
    "keras_model.summary()\n",
    "\n",
    "keras_model.compile(\n",
    "    optimizer=tf.keras.optimizers.SGD(0.02),\n",
    "    loss=tf.keras.losses.SparseCategoricalCrossentropy(from_logits=True),\n",
    "    metrics=[tf.keras.metrics.SparseCategoricalAccuracy()],\n",
    ")\n",
    "\n",
    "# Train loop\n",
    "history = keras_model.fit(\n",
    "    train_images,\n",
    "    train_labels0,\n",
    "    batch_size=100,\n",
    "    epochs=20,\n",
    "    validation_data=(test_images, test_labels0),\n",
    ")\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 59,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Key: loss\n",
      "Key: sparse_categorical_accuracy\n",
      "Key: val_loss\n",
      "Key: val_sparse_categorical_accuracy\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAjcAAAHHCAYAAABDUnkqAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjguMywgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy/H5lhTAAAACXBIWXMAAA9hAAAPYQGoP6dpAABffUlEQVR4nO3deXhTVcIG8Pdm777QHQoFRBZlURQEx1GUXSuurCOCiKPCuCDj8s0oMI7iyuDu6IjoiIo7jiJQyiY7gmzK1kKh0J3SpmuS5p7vjzShoVvSJk1z8/6eJ0+Sm3Nvzuml9vUs90pCCAEiIiIihVD5ugJEREREnsRwQ0RERIrCcENERESKwnBDREREisJwQ0RERIrCcENERESKwnBDREREisJwQ0RERIrCcENERESKwnBDRB4jSRLmz5/v9n5ZWVmQJAlLly71eJ2IKPAw3BApzNKlSyFJEiRJwubNm+t9LoRAcnIyJEnCTTfd5IMaesbKlSshSRKSkpIgy7Kvq0NE7QjDDZFCGQwGfPrpp/W2b9y4EadPn4Zer/dBrTxn2bJlSElJQW5uLtatW+fr6hBRO8JwQ6RQY8eOxZdffomamhqn7Z9++ikGDhyIhIQEH9Ws9SoqKrBixQrMmTMHl112GZYtW+brKjWqoqLC11UgCjgMN0QKNWnSJJw9exZpaWmObWazGV999RUmT57c4D4VFRV47LHHkJycDL1ej549e+KVV16BEMKpnMlkwqOPPorY2FiEhYXh5ptvxunTpxs85pkzZ3DPPfcgPj4eer0el1xyCZYsWdKqtn377beoqqrCnXfeiYkTJ+Kbb75BdXV1vXLV1dWYP38+Lr74YhgMBiQmJuK2225DZmamo4wsy3jttdfQt29fGAwGxMbGYvTo0fjll18AND0f6MI5RvPnz4ckSfj9998xefJkREVF4Q9/+AMAYP/+/Zg2bRq6desGg8GAhIQE3HPPPTh79myDP7MZM2YgKSkJer0eXbt2xQMPPACz2Yzjx49DkiT861//qrff1q1bIUkSPvvsM3d/pESKovF1BYjIO1JSUjBkyBB89tlnGDNmDADgp59+QmlpKSZOnIjXX3/dqbwQAjfffDPWr1+PGTNmYMCAAVi9ejX++te/4syZM05/TO+991588sknmDx5MoYOHYp169bhxhtvrFeH/Px8XHXVVZAkCbNnz0ZsbCx++uknzJgxA0ajEY888kiL2rZs2TIMGzYMCQkJmDhxIp588kn873//w5133ukoY7VacdNNNyE9PR0TJ07Eww8/jLKyMqSlpeHgwYPo3r07AGDGjBlYunQpxowZg3vvvRc1NTX4+eefsX37dlxxxRUtqt+dd96JHj164Pnnn3cEw7S0NBw/fhzTp09HQkICfvvtN7z33nv47bffsH37dkiSBADIycnBoEGDUFJSgvvuuw+9evXCmTNn8NVXX6GyshLdunXD1VdfjWXLluHRRx+t93MJCwvDuHHjWlRvIsUQRKQoH374oQAgdu3aJd58800RFhYmKisrhRBC3HnnnWLYsGFCCCG6dOkibrzxRsd+3333nQAg/vnPfzod74477hCSJImMjAwhhBB79+4VAMSDDz7oVG7y5MkCgJg3b55j24wZM0RiYqIoKipyKjtx4kQRERHhqNeJEycEAPHhhx822778/Hyh0WjE+++/79g2dOhQMW7cOKdyS5YsEQDEokWL6h1DlmUhhBDr1q0TAMRDDz3UaJmm6nZhe+fNmycAiEmTJtUra29rXZ999pkAIDZt2uTYNnXqVKFSqcSuXbsardO///1vAUAcOnTI8ZnZbBYxMTHi7rvvrrcfUaDhsBSRgo0fPx5VVVX44YcfUFZWhh9++KHRIamVK1dCrVbjoYcectr+2GOPQQiBn376yVEOQL1yF/bCCCHw9ddfIzU1FUIIFBUVOR6jRo1CaWkp9uzZ43abPv/8c6hUKtx+++2ObZMmTcJPP/2Ec+fOObZ9/fXXiImJwV/+8pd6x7D3knz99deQJAnz5s1rtExL3H///fW2BQUFOV5XV1ejqKgIV111FQA4fg6yLOO7775Dampqg71G9jqNHz8eBoPBaa7R6tWrUVRUhD/96U8trjeRUjDcEClYbGwshg8fjk8//RTffPMNrFYr7rjjjgbLnjx5EklJSQgLC3Pa3rt3b8fn9meVSuUY1rHr2bOn0/vCwkKUlJTgvffeQ2xsrNNj+vTpAICCggK32/TJJ59g0KBBOHv2LDIyMpCRkYHLLrsMZrMZX375paNcZmYmevbsCY2m8dH3zMxMJCUlITo62u16NKVr1671thUXF+Phhx9GfHw8goKCEBsb6yhXWloKwPYzMxqNuPTSS5s8fmRkJFJTU51Wwy1btgwdO3bE9ddf78GWEPknzrkhUrjJkydj5syZyMvLw5gxYxAZGdkm32u/9syf/vQn3H333Q2W6devn1vHPHbsGHbt2gUA6NGjR73Ply1bhvvuu8/NmjatsR4cq9Xa6D51e2nsxo8fj61bt+Kvf/0rBgwYgNDQUMiyjNGjR7foOj1Tp07Fl19+ia1bt6Jv3774/vvv8eCDD0Kl4v+zEjHcECncrbfeij//+c/Yvn07li9f3mi5Ll26YO3atSgrK3PqvTl8+LDjc/uzLMuOnhG7I0eOOB3PvpLKarVi+PDhHmnLsmXLoNVq8d///hdqtdrps82bN+P111/HqVOn0LlzZ3Tv3h07duyAxWKBVqtt8Hjdu3fH6tWrUVxc3GjvTVRUFACgpKTEabu9J8sV586dQ3p6OhYsWIBnnnnGsf3YsWNO5WJjYxEeHo6DBw82e8zRo0cjNjYWy5Ytw+DBg1FZWYm77rrL5ToRKRkjPpHChYaG4p133sH8+fORmpraaLmxY8fCarXizTffdNr+r3/9C5IkOVZc2Z8vXG21ePFip/dqtRq33347vv766wb/WBcWFrrdlmXLluGaa67BhAkTcMcddzg9/vrXvwKAYxn07bffjqKionrtAeBYwXT77bdDCIEFCxY0WiY8PBwxMTHYtGmT0+dvv/22y/W2BzFxwZL6C39mKpUKt9xyC/73v/85lqI3VCcA0Gg0mDRpEr744gssXboUffv2dbsnjEip2HNDFAAaGxaqKzU1FcOGDcPf/vY3ZGVloX///lizZg1WrFiBRx55xDHHZsCAAZg0aRLefvttlJaWYujQoUhPT0dGRka9Y77wwgtYv349Bg8ejJkzZ6JPnz4oLi7Gnj17sHbtWhQXF7vchh07diAjIwOzZ89u8POOHTvi8ssvx7Jly/DEE09g6tSp+PjjjzFnzhzs3LkT11xzDSoqKrB27Vo8+OCDGDduHIYNG4a77roLr7/+Oo4dO+YYIvr5558xbNgwx3fde++9eOGFF3DvvffiiiuuwKZNm3D06FGX6x4eHo4//vGPeOmll2CxWNCxY0esWbMGJ06cqFf2+eefx5o1a3DttdfivvvuQ+/evZGbm4svv/wSmzdvdhpWnDp1Kl5//XWsX78eL774osv1IVI83y3UIiJvqLsUvCkXLgUXQoiysjLx6KOPiqSkJKHVakWPHj3Eyy+/7FiCbFdVVSUeeugh0aFDBxESEiJSU1NFdnZ2vaXRQtiWbs+aNUskJycLrVYrEhISxA033CDee+89RxlXloL/5S9/EQBEZmZmo2Xmz58vAIh9+/YJIWzLr//2t7+Jrl27Or77jjvucDpGTU2NePnll0WvXr2ETqcTsbGxYsyYMWL37t2OMpWVlWLGjBkiIiJChIWFifHjx4uCgoJGl4IXFhbWq9vp06fFrbfeKiIjI0VERIS48847RU5OToM/s5MnT4qpU6eK2NhYodfrRbdu3cSsWbOEyWSqd9xLLrlEqFQqcfr06UZ/LkSBRhLign5SIiLyG5dddhmio6ORnp7u66oQtRucc0NE5Kd++eUX7N27F1OnTvV1VYjaFfbcEBH5mYMHD2L37t149dVXUVRUhOPHj8NgMPi6WkTtBntuiIj8zFdffYXp06fDYrHgs88+Y7AhugB7boiIiEhR2HNDREREisJwQ0RERIoScBfxk2UZOTk5CAsLa9Vdf4mIiKjtCCFQVlaGpKSkZu+hFnDhJicnB8nJyb6uBhEREbVAdnY2OnXq1GSZgAs39hsCZmdnIzw83KPHtlgsWLNmDUaOHNnojfqUgm1VrkBqL9uqXIHU3kBpq9FoRHJystONfRsTcOHGPhQVHh7ulXATHByM8PBwRf8DA9hWJQuk9rKtyhVI7Q2ktgJwaUoJJxQTERGRojDcEBERkaIw3BAREZGiMNwQERGRojDcEBERkaIw3BAREZGiMNwQERGRojDcEBERkaIw3BAREZGiMNwQERGRojDcEBERkaIw3BAREZGiBNyNM4mIiMiDhADkGqCmGqgx2Z4lFRCe5LMqMdwQERG1V0IAstUWHuQaQLacf2+12J7NVQirOg3kHwSkuuVrP5et58vaH1azcxixVDu9FzXVEBbbQ679TDjKmSBZqyHVmCBZTVBZqyEJ2anapbFXImLWWh/90BhuiIgoUAlh+4NuKgfM9kdFA+/L6n9mqTwfHIQAhLX2tWx7LWRAls+/b+IzUfuZ/dlW1raPStQ02wwtgOsB4LDnfjRS7QMA1G7uaxJaZJeYEOG56riN4YaIiNonIWp7EqoAS+2jpvr8a0uV82d13qtMFeh36hDU362wbTOV2cLJhSFFWH3dSkeIkJos5axGqGCFGhaoYYUKNVA7HlahggUaWFG3jBo1UKFGaGzPUKMaOlRDB5PQwgQtTNDBBC2qha72fe1DaFENHWSVHjUqHawqPaxqPWSVHrJaD1mtg1AbIDR6QK2DRqNBz4QwXOqNH5aLGG6IiKh59nkVlqo6gcM+lFHtvL3GdD6IOIYyGtluqax9XVn7/oLAAtGi6qoBdAWAs66Vr4IBlVIQqmBABQyoEAaUCwPKhR5G2YBKGFAOAyqFAeUIQpXQwwI1BCRYoYIMFWRIkKFyem+FylZGNPT5+TL2zzQaDQxaLTRaDdRqLVQaLVRqLSS1BmqtFmq1FhqNFlqNGjqNClq1ChoVkJN9Cj26d4VBp4FOrYZWI0GnVkGnUUGntpWzl9drVNCoJUSrVdBqVNCqVNBqJGjVF7xWq6BVS7XfIUGS3IlfvsVwQ0SkZJZqW69FxVlEVh6HdGITYK0Eqo2AyWj7rLq0zus6203ldUJMlW3IxEdqoEI19KgSOlQLHapqex2qoIdJaFEFve290Dl6JCpFbVCxhxUE1YaT2rAibM+V0EO4sXhYq5ag06gQrNcgWKdGsM7+bHuE6zQI0qkRotcgSKtGiF6NIJ0GIY4yF+ynt70O0qqhVrkfICwWC1auzMLY0T2h1Wrd3l+JGG6IiDzFagGqSoDqEqDqnO1RY6qdW9HIwzEXw/4Q5+dl1Ct74fua2iEWY51QYoSotocTIySrGYBtXsa1AHDEM001QQcztDBJOscwhi14aBwBwz7UYf+8+sLXtWHEhNqwInTnQwp0MNUJMTUN/LnSaVS1gUGDEL3zs0EjoSgvB927dkGYToMOGhV0ajX0WpWjR0OvOf+sr+0Jqbtdp1ZBr1U7l1eroGpBAKG2xXBDRFSXELa5GFUl5wNK3bDS4PYS28Nc5suaOzT0p7dMBKEcQSgTQShDMMpEMMoRBGPt+/ILtyMYFcJg6xmpnYdRXRtWzNC43NOhrw0FBq0aBq3a8dr+HKJXI0SnQaze1tMRUtubEVLbKxKi0yC4towjwNRu06obr4OtN+M0xo7tzd6MAMRwQ0TKYg8n9p4Mx3Pp+eGX2m3qqhIMPnUM6qVvAKbS86FFbn6FSpMMEYAhEgiKBLTBEJBQIyRYZMAiA2YrYJYlmK2AySpgssL2qBGotgpU1wDW2jkYAhJkYZ+rIUE45mfYtlXA4BRMjAhyel+lCoZFEwKtRgO5xozwkGBbuNDW9laoVbWvVdBpbKEjRqNCx9qeDENtOXef7cf0p3kapBwMN0TUvlhrantESs4/m2qDSb3AUue5utRWzlTm8twQFYAEADDW/0yotJANUajRR8CijYBZG4FqTRiqNOGoVIWhTBWGcikUpQhFiQjFOTkIZ+UQnK0JQrlFoMpsRWWFFecqLThXaYZVdn9ibLhBg5hQPaJDdIgO0aFDqB4dQnToEKpzbAvWaRy9I/ZQYQ8sOrUKmtreDVtPxkqMHXsNezJI8RhuiMjzZLm2J6SkzvBNE68d20o8NrQjVBrIujDUaMNg1oSiWhWCCikE5VIwjHIwzskGnK3R43S5hAptNIpqQlBQE4R8SxDOiRBUQQ9UutvrUFn7aFiYQVMbTmyB5XxQ0SOmNrB0CNGjQ6gOUcE66DS8Qw5RSzDcEJHrqkuBkmyg9DRQmg2UnAKMZ4DKs84TaauNaOkSXgd9eO3QTgSsunCYNWGoVoegUgpBOYJhFEEokYNwtkaPIosB+WY98kw65FRrcbpSi1Kr1vVwUtVIFTQqx4oX26oWDYIveB1UuwKmsXLhQVrEhOoRFaKFXuPu5dCIqCUYbojIRpaB8nxIZ08g6dx2qLZlAGW5tSGmNtCYSt07pjbENu/EPv8kKAowREI2RKBKHY4yKQQlIgTFcggKLUHItxiQY9LjdJUehRU1KCo342yxCdWWli1B1mtU6BCiQ1TtEE5UsO05MliL6BAdwvVqHDnwK677wxCEBek9siyXiHyP4YYoUNSY6vS4ZNueS0/bel9Ks4HSM4BsgQbAlQCQ1chxgqKByGQgwv7oBITEoloThhIRgrNycG1QCUJhpYyicjOKyk04W27G2RITisrNOFdphmiwY8dS+yiv94lWLTnCieM5RIvo4PrhJSpEh+hgHYJ0TfeUWCwWSNkCl3eO5DwUIgVhuCHyd7IVqCgCyvOAsjqPuu+NZ4Dy/OaPJakhwhJRbA1GVNd+kCK7oNyQgGJtPPKkWJyxdkBOlQoFZSYUGE0oOFGNgjJbcKmyWGELJfWDSYNfJQHRwbY5Jx1C9IgJs02WjQnVISZUb5s8G6pDTIge0aE6hOjUXHlDRC5huCFqr6w1QEVhbUjJtw0Rldc+131fXuD6/XG0wUBEMuSITqgKToJRl4AiTRxyEYNT1hicMIUh12jBsdMFMB814Gy5GTWOVT7nah+N02tUiAm1BZWY2smy9qByPrTYwkxUsNaxkoeIyJMYbojamizbJuCW5QDGXNtzQz0uFYWuX+5eUgEhsUBYAuTQeFTqYnBO3QEFciSyayJwwhyBY6YoHC/XofCcGcVnGhoWKqt9ALbLwJlsrySgQ4gOsWEGxIXpbY9wPeLs78P1tkATapuzwt4VIvI1hhsiTzJX2npUjDm1PSy55wOMsfZ9WR4gW1w7nqQCQuKAsATHQ4TGo0wXgwI5CqetEThRFYqj5QacLDEj+1wlck5WN3JNFXPtw0ajkhATag8qekd46RCiQfaRgxgzbCiSokLRIVTX5JVgiYjaG4YbIlcI2RZKqgovCC0XBJlqV1cTSY6eFoQn1QaXRNtzaALKtDE4Yw1HVnUwss/ZQkt2cSWy86pw+lxlndVDFjQ0VKRTq9ApKggdo4KQHB2MjpFBiLX3uoQZEBeuR3SwrsF75FgsFqwsPIC+HSM4yZaI/BLDDdGFzBVA/m9A7j4gdy80OfuQWvA7VHtdndcSAoQn1oaVxNrXSc7bwhJQUSPhSH4ZjuSVIbOgHNnZlcgurkL2uUqUVecDaHwCsEoCEiOC0Kk2vCRHBSM5+vzruDA9b+5HRAGL4YYCW7URyNtfG2RqH0VHnea6SLUPIakghcbXBpY6vS3hSc5BRh9um6hSq8Yq40RRBQ7nleHI72U4nJePI/nHkF3cyJXjasWE6tApKrg2sDiHmMSIIF69loioEQw3FDgqi51DTO4+oDiz4bKhCUBifyCxP2riLsW63wsxbNwUaPVBjR5eCIE8YzUOnyzEkTxbj8zh2l4Zs7XhicFxYXr0TAhDj7gwdLb3vEQHo1NUEIJ1/PUkImoJ/teTlKm8AMjZ6xhaQu5+oPRUw2UjOgOJ/YDEAbWBpp+tV6aWsFhQlbkSUJ3/dTFWWxzh5UieEUfzynE4zwhjdcN3kw7RqXFxQhh6JYShZ3wYeiaEo1dCGKJCdJ5rMxERAWC4ISWoMQHHNwJnfjnfI1OW23DZ6G6OHhkk9gcS+gMhHZo8fGGZCb8USvhtzVEcK6jAkbwy5JRWN1hWrZLQLSYEPe1BpjbEdIwM4hwYIqI2wnBD/kkI4PQvwP7PgYNf227WWJekAmIuviDI9AUMEc0eWpYFDuaUYt3hAqw7XID9p0sBqIGMLKdyiREG9EwIOx9k4sPRPS6EN0ckIvIxhhvyL+dOAvu/APZ95jxfJiwR6H79+aGlhEsBXYjLhy031WDzsSKsP1yAdUcKUFhmcvo8OUTgD32S0ScpAj0TwtEzPgwRwVwmTUTUHjHcUPtXXQr8vgLY9zlwcsv57dpgoHcq0H8i0PVaQOVej8mps5VIP5yPdYcLsON4sdOk3xCdGtf0iMX1veLwh+5R2PVzOsaO7cPrvhAR+QGGG2qfrDVA5jpbD82RlUCNfY6LBHT9I9B/ki3Y6ENdPqTFKmP3yXOO4aaMAucbPHaODsYNveNwfa84DOoa7RheslhcvJowERG1Cww31H4IYbvmzL7PgQNfARUF5z+L7WXroek7Hojo6PIhz1WYseFoAdIPFWDT0UKn1UxqlYQrU6JwQ694DOsVh+6xIbwvEhGRAjDckO8Zc4ADX9pCTcHv57cHxwB97wT6T7DNpXEheAghcCS/DOmHCrD+cAH2nDqHurdZigrWYljPOFzfOw7X9IhFRBCHmYiIlIbhhnzDXAEc+sE27HR8A4DaBKLWAz3H2IadLroBUDcfPoQQ2JxRhNW/5WH94UKcKXG+8m/vxHBc3ysW1/eKx4DkSKi5JJuISNEYbqjtyFYg62dbD83v3wOWivOfdR5iG3bqcwsQFOnyIfNKq/G3bw8g/fD5ISy9RoU/XBSDYb1s82eSIhu/qjARESkPww15nxDA1jeAHe8CxjPnt0d1tfXQ9BsPRHd185ACX/ySjX/+cAhlphro1CrcPrATRvSJw5BuMQjS8VozRESBiuGGvEsIYM3fgW1v2t4bIoBLb7eFmk5XujSP5kKnz1XiqW8O4OdjRQCA/smRePmOfrg4PsyTNSciIj/FcEPeIwSQ/o/zwWbU88CV9wIafYsOJ8sCy3aewgsrD6HCbIVeo8JjIy/GjD904zwaIiJyYLgh79n4IrB5ke312FeAQTNbfKiTZyvwxNf7sf14MQDgypQovHh7P3SLdf06N0REFBgYbsg7Nr0CbFhoez3q+RYHG6sssHRrFl5efRjVFhlBWjWeGN0TU4ek8EaURETUIIYb8rwtrwPrnrW9Hr4AGDKrRYfJLCzH41/tx+6TtptiDu3eAS/c1g+dOwR7qqZERKRADDfkWdvfBdKetr0e9nfgD4+4fYgaq4z/bD6BRWlHYa6REarX4KmxvTB5UGdeQZiIiJrFcEOes+s/wKonbK//+Dhw7V/dPsSRvDI8/tU+7DtdajvMxbFYeFtfdOS1aoiIyEUMN+QZez4GfnzM9vrqR4Bh/+fW7harjHc3ZOL1dcdgsQqEGTR45qY+uGNgJ/bWEBGRWxhuqPX2fgZ8/5Dt9VWzgOHz3bp+zcEzpXj8q/34PdcIABjeOw7P3doX8eEGL1SWiIiUjuGGWkX67WtgxYMABHDlTGDUcy4HG1ONFW+uy8A7GzJRIwtEBmux4OZLcHP/JPbWEBFRizHcUIslntsJ9d53ACEDA6cBY15yOdjszS7B41/tw9H8cgDA2L4JWHDzpYgNa9kF/oiIiOwYbqhFpCMrcUXWO5BgBQZMAW78F6BSNbtftcWKf6Udxfs/H4csgJhQHf4x7lKM7ZvYBrUmIqJAwHBD7ju6BupvZkCCFfKld0B18xsuBZtfsorx+Ff7cbzIdjfwWwYk4ZnUSxAdovN2jYmIKIAw3JB7MtKB5X+CJFtwJnIQ4lLfhErV/B24X1l9BG9tyIAQQFyYHs/d2hcj+sS3QYWJiCjQMNyQ645vBD6fDFhNkHveiN1Bt2OMqvl/QlszivDm+gwAwJ0DO+HvN/VBRJDW27UlIqIA1fxYAhEAnNwKfDYRqKkGLh4N663vQ0jNBxtZFnj+p0MAgKlDuuDlO/sz2BARkVex54aal70TWHYnYKkEut8A3PkRgOaHogDgf/tzcPCMEaF6DR66oYd360lERAT23FBzzuwGPrkdMJcDXf8ITFwGaF27uJ6pxoqXVh0BANx/bTfEhHKZNxEReR/DDTUudx/w31sBkxHocjUw6XNA6/o9nv677STOlFQhPlyPGX/o5sWKEhERncdwQw3LOwh8PA6oLgWSBwOTlwO6EJd3L6204I11tknEj43oiSCda8NYRERErcVwQ/UVHLYFm6pzQMeBwJQvAX2YW4d4e0MGSqssuDg+FLcP7OSlihIREdXHcEPOijKAj28GKouAxP7An74GDBFuHeL0uUp8uDULAPDUmN5Qq3ifKCIiajsMN3Re8XHgo1SgPB+IvxS46zsgKMrtw7y65ijMNTKGdOuA63rGer6eRERETWC4IZtzJ4GPbgbKcoDYXsDUFUBwtNuHOXimFN/+egYA8NTYXry7NxERtTmGG7L538NAaTbQoQcw9XsgJMbtQwghsLD2gn03909Cv06RHq4kERFR8xhuCKgoAk5stL2evBwIa9k9nzYdK8KWjLPQqVX466ieHqwgERGR63webt566y2kpKTAYDBg8ODB2LlzZ5PlFy9ejJ49eyIoKAjJycl49NFHUV1d3Ua1VajDPwBCBhIHAB26t+gQVllg4crzt1lIjg72YAWJiIhc59Nws3z5csyZMwfz5s3Dnj170L9/f4waNQoFBQUNlv/000/x5JNPYt68eTh06BA++OADLF++HP/3f//XxjVXmN+/tz33ubnFh/j21zM4nFeGcIMGs6+/yEMVIyIicp9Pw82iRYswc+ZMTJ8+HX369MG7776L4OBgLFmypMHyW7duxdVXX43JkycjJSUFI0eOxKRJk5rt7aEmVJ07PyTVe1yLDlFtseLVNbbbLMwadhEig3Weqh0REZHbfHbjTLPZjN27d+Opp55ybFOpVBg+fDi2bdvW4D5Dhw7FJ598gp07d2LQoEE4fvw4Vq5cibvuuqvR7zGZTDCZTI73RqMRAGCxWGCxWDzUGjiOWffZH0i//wiNXAMR2xs1EV0AF+tet61Ltp1Gbmk1kiIMmHJlR79qvyv88by2RiC1l21VrkBqb6C01Z32SUII4cW6NConJwcdO3bE1q1bMWTIEMf2xx9/HBs3bsSOHTsa3O/111/H3LlzIYRATU0N7r//frzzzjuNfs/8+fOxYMGCets//fRTBAdzXsig4/9CYumvOJxwC44k3ub2/uUW4Nlf1ai2SvjTRVZcGeuTf05ERKRwlZWVmDx5MkpLSxEeHt5kWZ/13LTEhg0b8Pzzz+Ptt9/G4MGDkZGRgYcffhjPPvssnn766Qb3eeqppzBnzhzHe6PRiOTkZIwcObLZH467LBYL0tLSMGLECGi1Wo8e2ytMZdDsnwkA6J46B93j+ri8q72th9UpqLaeRu+EMDx911VQKfBqxH53XlspkNrLtipXILU3UNpqH3lxhc/CTUxMDNRqNfLz85225+fnIyEhocF9nn76adx111249957AQB9+/ZFRUUF7rvvPvztb3+DSlV/CpFer4der6+3XavVeu0fgTeP7VGH1wFWExDdHdqkfoCbF9wrqgY+32+7YN/fbuwDvV7Zc2385rx6SCC1l21VrkBqr9Lb6k7bfDahWKfTYeDAgUhPT3dsk2UZ6enpTsNUdVVWVtYLMGq17W7TPhpd82+H6qySasGVhH84pYLFKvDHi2Pxhx7uX/SPiIjIG3w6LDVnzhzcfffduOKKKzBo0CAsXrwYFRUVmD59OgBg6tSp6NixIxYuXAgASE1NxaJFi3DZZZc5hqWefvpppKamOkIOuchcCRxLs73u7f4S8H2nS/HrWRUkCXhydC8PV46IiKjlfBpuJkyYgMLCQjzzzDPIy8vDgAEDsGrVKsTH266Qe+rUKaeemr///e+QJAl///vfcebMGcTGxiI1NRXPPfecr5rgvzLTAUslENEZSLrMrV2FEHhx9VEAwC0DktAnybNzl4iIiFrD5xOKZ8+ejdmzZzf42YYNG5zeazQazJs3D/PmzWuDmimc/cJ9vVPdHpJKP1SAXVnnoJEEHr2BF+wjIqL2xee3XyAfqDEBR1fZXrt5VeIaq4wXVh0GAFybKJAYYfB07YiIiFqF4SYQHd8ImIxAaALQaZBbu365+zQyCsoRFazFiI6ylypIRETUcgw3gejQCttz75uABpbPN6bSXINFaba5Ng9e1w1BPh/UJCIiqo/hJtBYa4DDK22v3Vwl9Z+fT6CwzITO0cGYfGWyFypHRETUegw3gebkZqCqGAiKBrpc7fJuhWUm/HtjJgDgr6N6QqfhPx0iImqf+Bcq0NhXSfW6EVC7Pq70WvpRVJit6N8pAjf2TfRS5YiIiFqP4SaQyDJw+Afb6z7jXN4ts7Acn+3MBgA8Oaa3Iu8fRUREysFwE0iydwDl+YA+Auh6rcu7vbTqMKyywA294jCkewcvVpCIiKj1GG4Cif1eUj1HAxrXbnL5S1YxVv+WD5UEPDmGt1kgIqL2j+EmUAgBHPqf7bWLq6SEEHh+5SEAwIQrk9EjPsxbtSMiIvIYhptAkbMHKM0GtCHARTe4tMvq3/Kw51QJgrRqPDL8Yi9XkIiIyDMYbgKFfZVUjxGANqjZ4harjBdXHQEAzLymK+LDeZsFIiLyDww3gUCI8/NtXLyX1Gc7T+FEUQU6hOhw37XdvVg5IiIiz2K4CQT5vwHFxwG1HugxstniZdUWvLb2GADgkeE9EKrnfRaIiMh/MNwEAnuvzUU3APrmJwW/t+k4zlaY0S0mBBMHdfZy5YiIiDyL4SYQ2OfbuLBKKt9Yjfd/Pg4AeHx0L2jV/CdCRET+hX+5lK7oGFB4CFBpbNe3aca/0o6i2iJjYJcojLokvg0qSERE5FkMN0r3+wrbc9drgaCoJosezS/DF7/YbrPwf2N7QZJ4mwUiIvI/DDdK58YqqRd+OgxZAKMvScDALtFerhgREZF3MNwo2bksIHcfIKmAXjc1WXRrZhHWHS6AWiXh8dE926Z+REREXsBwo2T22y10uRoIiWmy6HubbJOIJw/qjG6xod6uGRERkdcw3CiZG6ukfssxAgBuu7yjN2tERETkdQw3SmXMAU7vtL3u3fSQVGmVBYVlJgBA9zj22hARkX9juFGqQz/YnjsNAsKTmiyaWVgOAIgP1yPcoPV2zYiIiLyK4Uap3FgllVFgCzcXsdeGiIgUgOFGiSqKgJNbbK97pzZbPNMebjiRmIiIFIDhRokO/wAIGUjsD0SlNFucPTdERKQkDDdK5MYqKQDIqJ1z0509N0REpAAMN0pTdQ44sdH2us+4ZotXW6zILq4EwJ4bIiJSBoYbpTmyCpBrgNjeQEyPZoufKKqALIAwgwaxYfo2qCAREZF3MdwojRurpADn+Ta8USYRESkBw42SmMqAjHTba1fn23ClFBERKQzDjZIcWwNYTUB0NyD+Epd2sV/Aj/NtiIhIKRhulKTuKikXh5i4DJyIiJSG4UYpLFXAsTTbaxfn21hlgeNFFQAYboiISDkYbpQiIx2wVAARyUDS5S7tcvpcJcw1MnQaFTpFBXu5gkRERG2D4UYp7Kukeqe6PSTVLSYEahVXShERkTIw3ChBjdl2fRvA5VVSwPlw051DUkREpCAMN0pwYiNgKgVC44HkwS7vxmXgRESkRAw3SvD7Cttzr5sAleunNIPLwImISIEYbvydtQY4/KPttYurpABACMFl4EREpEgMN/7u5BagqhgIiga6/MHl3QrLTSirroFKArrGhHixgkRERG2L4cbf2VdJ9RoLqDUu72bvtUmODoZBq/ZGzYiIiHyC4cafyTJw6Afb6z63uLVrJicTExGRQjHc+LPTO4HyPEAfAXS91q1dOd+GiIiUiuHGn9lXSfUcDWh0bu1qXynFa9wQEZHSMNz4KyGAQ/+zvXbjwn12jgv4cViKiIgUhuHGX+XsAUqzAW0IcNENbu1qrLYg32gCwGEpIiJSHoYbf/V77SqpHiMAbZBbu9onE8eG6RERpPV0zYiIiHyK4cYfCXF+CbgbF+6z420XiIhIyRhu/FH+b0DxcUCtB3qMdHt33naBiIiUjOHGH9l7bS66AdCHub17ZkGFbXeGGyIiUiCGG39kn2/TglVSAJDJnhsiIlIwhht/U3QMKDwEqDS269u4yVRjxcmz7LkhIiLlYrjxN/YL93W9FgiKcnv3rKJKyAII02sQF6b3cOWIiIh8j+HG37RilRRQ5+J9caGQJMlTtSIiImo3GG78ybksIHcfIKmAXje16BC8MjERESkdw40/sd9uocvVQEhMiw7BZeBERKR0DDf+pJWrpADeDZyIiJSP4cZfGHOA0zttr3u3bEjKKgscZ88NEREpHMONvzjyk+250yAgPKlFh8gpqYKpRoZOrUJylHv3oyIiIvIXDDf+ouio7bnLkBYfwj4k1TUmBBo1Tz0RESkT/8L5i9LTtueI5BYfgvNtiIgoEDDc+At7uAnv2OJD1L3GDRERkVIx3PgL4xnbc0SnFh+Cy8CJiCgQMNz4A0s1UFFoe93CcCOEqHMBvxBP1YyIiKjdYbjxB/ZeG01Qi+4nBQBF5WaUVlkgSbw6MRERKRvDjT+oOyTVwvtB2XttOkUFwaBVe6pmRERE7Q7DjT9wrJRqxWRi+3wb9toQEZHC+TzcvPXWW0hJSYHBYMDgwYOxc+fOJsuXlJRg1qxZSExMhF6vx8UXX4yVK1e2UW19pLT1k4kzuQyciIgChMaXX758+XLMmTMH7777LgYPHozFixdj1KhROHLkCOLi4uqVN5vNGDFiBOLi4vDVV1+hY8eOOHnyJCIjI9u+8m3JaF8G3opww5VSREQUIHwabhYtWoSZM2di+vTpAIB3330XP/74I5YsWYInn3yyXvklS5aguLgYW7duhVarBQCkpKS0ZZV9wxPDUuy5ISKiAOGzcGM2m7F792489dRTjm0qlQrDhw/Htm3bGtzn+++/x5AhQzBr1iysWLECsbGxmDx5Mp544gmo1Q1PkjWZTDCZTI73RqMRAGCxWGCxWDzYIjiO5+njakpPQwJQE5II0YJjl5tqkFtaDQDoEmXwSP281db2KJDaCgRWe9lW5Qqk9gZKW91pn8/CTVFREaxWK+Lj4522x8fH4/Dhww3uc/z4caxbtw5TpkzBypUrkZGRgQcffBAWiwXz5s1rcJ+FCxdiwYIF9bavWbMGwcHBrW9IA9LS0jx6vLFns6AFsHFvBsoPV7q9/8lyANAgTCuwZb1n6+bptrZngdRWILDay7YqVyC1V+ltrax0/e+fT4el3CXLMuLi4vDee+9BrVZj4MCBOHPmDF5++eVGw81TTz2FOXPmON4bjUYkJydj5MiRCA8P92j9LBYL0tLSMGLECMewWatVG6H91dbr8sebJgE69y/A9+2vOcCBg+jTKRpjx17pkWp5pa3tVCC1FQis9rKtyhVI7Q2UttpHXlzhs3ATExMDtVqN/Px8p+35+flISEhocJ/ExERotVqnIajevXsjLy8PZrMZOp2u3j56vR56vb7edq1W67V/BB49dnHtzycoCtqQyBYd4kRxFQDgorgwj7fZmz/H9iaQ2goEVnvZVuUKpPYqva3utM1nS8F1Oh0GDhyI9PR0xzZZlpGeno4hQ4Y0uM/VV1+NjIwMyLLs2Hb06FEkJiY2GGwUobT1K6U4mZiIiAKJT69zM2fOHLz//vv46KOPcOjQITzwwAOoqKhwrJ6aOnWq04TjBx54AMXFxXj44Ydx9OhR/Pjjj3j++ecxa9YsXzXB+4ytXynFa9wQEVEg8emcmwkTJqCwsBDPPPMM8vLyMGDAAKxatcoxyfjUqVNQqc7nr+TkZKxevRqPPvoo+vXrh44dO+Lhhx/GE0884asmeF8rL+BnrpFxstg2CYvhhoiIAoHPJxTPnj0bs2fPbvCzDRs21Ns2ZMgQbN++3cu1akccw1It67k5ebYCVlkgVK9BQrjBgxUjIiJqn3x++wVqhrF1PTf2+TbdY0MgtfCmm0RERP6E4aa9c1yduJXhhkNSREQUIBhu2jNZPt9z08JhqQzeU4qIiAIMw017VlkEWM0AJCA8qUWHcCwDj2W4ISKiwMBw057Zh6TCEgC1+xdmkmXBu4ETEVHAYbhpz1q5UupMSRWqLTK0agmdo71zHy0iIqL2huGmPXOslGrdfJuUDiHQqHmqiYgoMPAvXnvmWCmV3KLdeWViIiIKRAw37Vkrh6U434aIiAIRw0171tphKfbcEBFRAHI73Hz44Yf48ssv623/8ssv8dFHH3mkUlSrlfeVOn91YoYbIiIKHG6Hm4ULFyImJqbe9ri4ODz//PMeqRQBsFqAslzb63D3w83ZchPOVVogSQw3REQUWNwON6dOnULXrl3rbe/SpQtOnTrlkUoRaoONAFRaICTW7d3tvTYdI4MQpFN7uHJERETtl9vhJi4uDvv376+3fd++fejQoYNHKkWoMyTVEVC5PzWKt10gIqJA5fZfzUmTJuGhhx7C+vXrYbVaYbVasW7dOjz88MOYOHGiN+oYmBwrpTjfhoiIyB0ad3d49tlnkZWVhRtuuAEajW13WZYxdepUzrnxJKNn7gbOnhsiIgo0bocbnU6H5cuX45///Cf27t2LoKAg9O3bF126dPFG/QJXaeuWgR8vrADAcENERIHH7XBj16NHD/To0cOTdaG6WnEBvwpTDc6UVAHg3cCJiCjwuD3n5vbbb8eLL75Yb/tLL72EO++80yOVItQZlnL/1gv2XpsOITpEheg8WSsiIqJ2z+1ws2nTJowdO7be9jFjxmDTpk0eqRShVcNSGYVlAIDuHJIiIqIA5Ha4KS8vh05XvzdAq9XCaDR6pFIBz1wJVBXbXrdgWIqTiYmIKJC5HW769u2L5cuX19v++eefo0+fPh6pVMCz31NKFwYYItze3RFuON+GiIgCkNsTip9++mncdtttyMzMxPXXXw8ASE9Px6effoqvvvrK4xUMSKXZtueIjoAkub07e26IiCiQuR1uUlNT8d133+H555/HV199haCgIPTv3x/r1q1DdHS0N+oYeOzzbVowJGWxyjh5thIAww0REQWmFi0Fv/HGG3HjjTcCAIxGIz777DPMnTsXu3fvhtVq9WgFA5Kx5XcDP3m2AjWyQLBOjcQIg4crRkRE1P65f9OiWps2bcLdd9+NpKQkvPrqq7j++uuxfft2T9YtcDmGpdwPNxkFtmXg3WNDIbVgSIuIiMjfudVzk5eXh6VLl+KDDz6A0WjE+PHjYTKZ8N1333EysSe1YlgqkzfMJCKiAOdyz01qaip69uyJ/fv3Y/HixcjJycEbb7zhzboFrlYMS3EyMRERBTqXe25++uknPPTQQ3jggQd42wVvEuL8rRdaEW54N3AiIgpULvfcbN68GWVlZRg4cCAGDx6MN998E0VFRd6sW2CqOgdYbKudEJ7k1q6yLDgsRUREAc/lcHPVVVfh/fffR25uLv785z/j888/R1JSEmRZRlpaGsrKyrxZz8BhH5IKjgG0QW7tmmusRqXZCo1KQpcOwV6oHBERUfvn9mqpkJAQ3HPPPdi8eTMOHDiAxx57DC+88ALi4uJw8803e6OOgcUxJNXy2y6kxIRAq27xQjgiIiK/1qq/gD179sRLL72E06dP47PPPvNUnQKbPdyEt2IyMefbEBFRAPPI/96r1Wrccsst+P777z1xuMDmgZVS3eNCPFkjIiIiv8Kxi/amFcNSnExMRETEcNP+tOYCfo5hqTBP1oiIiMivMNy0N0Z7z02yW7udqzDjbIUZAIeliIgosDHctCeyFTDm2F67OSyVUTsk1TEyCMG6Ft0PlYiISBEYbtqT8gJArgEkFRCa4Nau5ycTc74NEREFNoab9sS+UiosCVC71/vCZeBEREQ2DDftSWm27bkVF/DjSikiIgp0DDftSStWSjHcEBER2TDctCctvIBfpbkGZ0qqADDcEBERMdy0J45hKffCzfHCCgBAVLAW0SE6T9eKiIjIrzDctCelLeu54ZWJiYiIzmO4aU+MLZtzw/k2RERE5zHctBc1JqA83/bazZ4bxzVuuAyciIiI4abdsF+ZWGMAgju4tSt7boiIiM5juGkv7HcDD+8ISJLLu9VYZWSdtU0oZrghIiJiuGk/HMvA3Ztvc7K4EharQJBWjaSIIC9UjIiIyL8w3LQXpS27G/j5e0qFQKVyvceHiIhIqRhu2ou6w1Ju4D2liIiInDHctBctHJbiNW6IiIicMdy0Fy29gB+XgRMRETlhuGkvHMNSrocbIQQyC7lSioiIqC6Gm/bAVAaYSm2v3RiWyjNWo9xUA7VKQpcOIV6qHBERkX9huGkP7ENShghAH+bybvbJxF06BEOn4akkIiICGG7ahxYMSQFcKUVERNQQhpv2wGi/xg1vmElERNRaDDftQQtXSjHcEBER1cdw0x608AJ+vMYNERFRfQw37YFjWMr1npuSSjOKys0AeI0bIiKiuhhu2oMWDEvZe20SIwwI0Wu8USsiIiK/xHDja0Kcv/WCG8NSnG9DRETUMIYbX6s8C9RU216HJ7m8WwZvu0BERNQghhtfs08mDo0HNHqXd2PPDRERUcMYbnythSulMrhSioiIqEEMN75mn2/jxgX8qi1WnD5XBYDhhoiI6EIMN75m77mJSHZ5l8zCcggBRAZr0SFE56WKERER+ad2EW7eeustpKSkwGAwYPDgwdi5c6dL+33++eeQJAm33HKLdyvoTS0Ylqp7TylJkrxRKyIiIr/l83CzfPlyzJkzB/PmzcOePXvQv39/jBo1CgUFBU3ul5WVhblz5+Kaa65po5p6ibEl17ipAMAhKSIioob4PNwsWrQIM2fOxPTp09GnTx+8++67CA4OxpIlSxrdx2q1YsqUKViwYAG6devWhrX1gpZcwI8rpYiIiBrl00vbms1m7N69G0899ZRjm0qlwvDhw7Ft27ZG9/vHP/6BuLg4zJgxAz///HOT32EymWAymRzvjUYjAMBiscBisbSyBc7sx3P5uHINNGU5kABYguMAF/c7ll8GAOgSbfB4G1zldlv9WCC1FQis9rKtyhVI7Q2UtrrTPp+Gm6KiIlitVsTHxzttj4+Px+HDhxvcZ/Pmzfjggw+wd+9el75j4cKFWLBgQb3ta9asQXBwsNt1dkVaWppL5QzmsxglZMiSGis37Qak5jvSrALILFQDkJB9cBdWZrSysq3kaluVIJDaCgRWe9lW5Qqk9iq9rZWVlS6X9aubEpWVleGuu+7C+++/j5iYGJf2eeqppzBnzhzHe6PRiOTkZIwcORLh4eEerZ/FYkFaWhpGjBgBrVbbbHnp9E7gN0AK74ixN97k0ndkna2AdfsWGLQqTLllDFQq30wodret/iyQ2goEVnvZVuUKpPYGSlvtIy+u8Gm4iYmJgVqtRn5+vtP2/Px8JCQk1CufmZmJrKwspKamOrbJsgwA0Gg0OHLkCLp37+60j16vh15f/8q/Wq3Wa/8IXD52eS4AQIro5HJdsoptQ2zdYkKh1/t+Gbg3f47tTSC1FQis9rKtyhVI7VV6W91pm08nFOt0OgwcOBDp6emObbIsIz09HUOGDKlXvlevXjhw4AD27t3reNx8880YNmwY9u7di+Rk168V0y60YKUUb7tARETUNJ8PS82ZMwd33303rrjiCgwaNAiLFy9GRUUFpk+fDgCYOnUqOnbsiIULF8JgMODSSy912j8yMhIA6m33C44L+PFu4ERERJ7i83AzYcIEFBYW4plnnkFeXh4GDBiAVatWOSYZnzp1CiqVz1ese4d9Gbg7F/DjPaWIiIia5PNwAwCzZ8/G7NmzG/xsw4YNTe67dOlSz1eorRjdu/WCEALH2XNDRETUJIV2ifgJN4elCspMKDPVQK2S0KWDd5axExER+TuGG1+xVAGVZ22vXRyWss+36RwdDL1G7a2aERER+TWGG18x5tietSFAUJRLu9jDTfdYDkkRERE1huHGV0qzbc8RHQEX7+zNlVJERETNY7jxlZaslGK4ISIiahbDja+05AJ+XAZORETULIYbX3EMS7kWbkqrLCgss916oXtsiLdqRURE5PcYbnzFzWEp+5BUQrgBYQbl3juEiIiotRhufMXNYalMDkkRERG5hOHGF4SocwE/18LNwTOlAICL48O8VSsiIiJFYLjxhepSwGzriXF1WGrniWIAwJUprl0Th4iIKFAx3PiCfUgqKBrQNX8bhXMVZhzOKwMADOoa7c2aERER+T2GG19w855Su7JsvTYXxYWiQ6jeW7UiIiJSBIYbX7CHm3DX5tvsqB2SGsxeGyIiomYx3PiCmyul7PNtOCRFRETUPIYbX3BjWMpYbcFvObaVUoO7dvBmrYiIiBSB4cYX7Bfwi0hutujurHOQBdClQzASIgxerhgREZH/Y7jxBaN9zk3zPTecb0NEROQehpu2Jst1em5cCTdnAQCDOCRFRETkEoabtlZRCMgWQFIBYYlNFq001+DAaft8G/bcEBERuYLhpq3Zh6RCEwB10zfA3HOyBDWyQFKEAZ2igtqgckRERP6P4aatubFSamftkNTgbh0gSZI3a0VERKQYDDdtrdT1a9xs5/VtiIiI3MZw09ZKXVspVW2xYm92CQDOtyEiInIHw01bs8+5aabnZl92Ccw1MmJC9egaE9IGFSMiIlIGhpu25uKwlP2WC4O7RXO+DRERkRsYbtqai8NSvHgfERFRyzDctKUaM1Ceb3vdRM+NxSpj98lzAHg/KSIiIncx3LSlslwAAlDrgeCYRosdOFOKKosVkcFa9IgLbbv6ERERKQDDTVtyDEklAarGf/T2+TaDUqKhUnG+DRERkTsYbtqS0bXJxDuO2+8nxfk2RERE7mK4aUulzS8Dt8oCv2TZ5ttc1Y3zbYiIiNzFcNOWXFgpdSjXiDJTDcL0GvRODG+jihERESkHw01bcgxLNR5u7EvAr0iJgprzbYiIiNzGcNOWHBfwS260yPn5NhySIiIiagmGm7ZUmm17bmRYSpYFdmWdvzIxERERuY/hpq2YK4DqEtvrRoaljhWU41ylBUFaNfp2jGi7uhERESkIw01bsQ9J6cMBQ8PBZecJ25DUwC5R0Kp5aoiIiFqCf0HbSjNDUgCwnfeTIiIiajWGm7bSzEopIQR2HK+9MjHDDRERUYsx3LSV0qavTnyiqAJF5SboNCr0T45su3oREREpDMNNW3FcwK/hcGO/vs2A5EgYtOq2qhUREZHiMNy0FWPTt16w3yzzKg5JERERtQrDTVspbXzOjW2+DS/eR0RE5AkMN21BiCbvK3X6XBVySquhUUm4vEtk29aNiIhIYRhu2kLVOaCmyva6gXBjn2/Tt1MEgnWatqwZERGR4jDctAV7r01ILKA11PvYfvG+wRySIiIiajWGm7bQxJAUcL7nhhfvIyIiaj2Gm7ZgbPwaN3ml1Th5thIqCRiYEtXGFSMiIlIehpu2UNr4MvAdtUNSfZLCEW7QtmWtiIiIFInhpi00MSy10zEkxfk2REREnsBw0xaaGJayz7fh/aSIiIg8g+GmLTQyLFVUbkJGQTkAYFAKww0REZEnMNx4m2wFjDm21xcMS+2q7bXpGR+GqBBdW9eMiIhIkRhuvK08HxBWQFIDYQlOHzmWgHdjrw0REZGnMNx4m2MycRKgcr7bN+fbEBEReR7Djbc1slKqtNKCw3lGAAw3REREnsRw422NrJTalVUMIYBuMSGIC6t/SwYiIiJqGYYbb3OslHLuudmZxfk2RERE3sBw422OYSnnnpsdx21XJuaQFBERkWcx3HhbA8NS5aYaHMyxzbfhlYmJiIg8i+HG2xoYltp98hysskCnqCAkRQb5qGJERETKxHDjTTUmoKLQ9rrOsJR9SIq9NkRERJ7HcONN9iEpTRAQfH5uzfmbZXK+DRERkacx3HhT3SEpSQIAVJmt2He6BABXShEREXkDw403ldb23NS5gN+v2edgsQrEh+vROTrYRxUjIiJSLoYbbzLae26SHZt2HLcPSXWAVNubQ0RERJ7DcONNDayU2sn7SREREXkVw403XTAsZaqxYs+pcwCAqzjfhoiIyCvaRbh56623kJKSAoPBgMGDB2Pnzp2Nln3//fdxzTXXICoqClFRURg+fHiT5X3qggv4HThdClONjA4hOnSPDfVhxYiIiJTL5+Fm+fLlmDNnDubNm4c9e/agf//+GDVqFAoKChosv2HDBkyaNAnr16/Htm3bkJycjJEjR+LMmTNtXHMXOIalbOFmR50hKc63ISIi8g6NryuwaNEizJw5E9OnTwcAvPvuu/jxxx+xZMkSPPnkk/XKL1u2zOn9f/7zH3z99ddIT0/H1KlT26TOLqk2AibbLRbsw1I7ON+GiMijZFmG2Wyut91isUCj0aC6uhpWq9UHNWs7SmqrTqeDStX6fhefhhuz2Yzdu3fjqaeecmxTqVQYPnw4tm3b5tIxKisrYbFYEB3dcGAwmUwwmUyO90ajLXBYLBZYLJZW1L4++/EsFgtQchJaAMIQiRqVHjXVJuyuvRP4wOQIj393W3Nqq8IFUluBwGov2+rfzGYzsrOzIctyvc+EEEhISMCpU6cU31OupLaqVCp07twZWq223mfu/Nv1abgpKiqC1WpFfHy80/b4+HgcPnzYpWM88cQTSEpKwvDhwxv8fOHChViwYEG97WvWrEFwsHeuM5OWloa40n0YAsCIMGxYuRIny4EKswZBaoHMPT/jhH//+3NIS0vzdRXaTCC1FQis9rKt/ik6OhpRUVGIjY31+z/qZAtphYWF2L17N4qLi+t9XllZ6fKxfD4s1RovvPACPv/8c2zYsAEGg6HBMk899RTmzJnjeG80Gh3zdMLDwz1aH4vFgrS0NIwYMQK6A4XAcSCsU2+MHTsW/9mcBRw4iiEXxeGmGy/z6Pf6Qt22NpSwlSSQ2goEVnvZVv9VU1ODEydOICkpqcH/lgshUFZWhrCwMMUHHyW1Va/XQ6VS4YorroBG4xxR7CMvrvBpuImJiYFarUZ+fr7T9vz8fCQkJDS57yuvvIIXXngBa9euRb9+/Rotp9frodfr623XarVe+wXXarXQVOQBAFSRyVBptdh9qgQAcFX3Dor4D4udN3+O7U0gtRUIrPayrf7HarVCkiTHH8ML2YeqJEnyyByO9kxJbdXr9ZAkCZIk1ft36s6/W5/+FHQ6HQYOHIj09HTHNlmWkZ6ejiFDhjS630svvYRnn30Wq1atwhVXXNEWVXVfnQv4WWVR52aZvBM4EZGn+HtPBTnz1Pn0ecSbM2cO3n//fXz00Uc4dOgQHnjgAVRUVDhWT02dOtVpwvGLL76Ip59+GkuWLEFKSgry8vKQl5eH8vJyXzWhYaXnb71wJK8MxuoahOjUuCTJs0NhREQUmFJSUrB48WJfV6Nd8vmcmwkTJqCwsBDPPPMM8vLyMGDAAKxatcoxyfjUqVNO3WzvvPMOzGYz7rjjDqfjzJs3D/Pnz2/LqjfNeP7qxDtOnAUADEyJhkbt8zxJREQ+ct1112HAgAEeCSW7du1CSEhI6yulQD4PNwAwe/ZszJ49u8HPNmzY4PQ+KyvL+xVqLSHO33ohoiN2braFm8G8vg0RETVBCAGr1VpvMm1DYmNjAaDBpfCBjt0I3lBZBFhNACSIsMQ6820YboiIAtW0adOwceNGvPbaa45Js0uXLoUkSfjpp58wcOBA6PV6bN68GZmZmRg3bhzi4+MRGhqKK6+8EmvXrnU63oXDUmq1Gv/5z39w6623Ijg4GD169MD333/fxq1sHxhuvECyz7cJjUdmsRlnK8zQa1To2ynCtxUjIlIoIQQqzTVOjyqztd42bzyEEC7V8bXXXsOQIUMwc+ZM5ObmIjc3F8nJyQCAJ598Ei+88AIOHTqEfv36oby8HGPHjkV6ejp+/fVXjB49GqmpqTh16lST37FgwQKMHz8e+/fvx9ixYzFlypQGrxmjdO1iWEpxjDm254iO2H7c9o/q8s5R0GvUPqwUEZFyVVms6PPMap989+//GIVgXfN/TiMiIqDT6RAcHOy43In9grX/+Mc/MGLECEfZ6Oho9O/f3/H+2Wefxbfffovvv/++0WkcgK13aNKkSQCA559/Hq+//jp27tyJ0aNHt6ht/oo9N14glZ2/G7j9flKDu3FIioiIGnbhZU3Ky8sxd+5c9O7dG5GRkQgNDcWhQ4ea7bmpe923kJAQhIeHN3ojaiVjz4031A5LifCO2LnHNpmYN8skIvKeIK0av/9jlOO9LMsoM5YhLDzM6xe2C9K2vlf+wlVPc+fORVpaGl555RVcdNFFCAoKwh133NHgTULruvBCd5IkBeSEY4YbL5Bql4Gf08Qh32iCVi3h8s5RPq4VEZFySZLkNDQkyzJqdGoE6zTt6qq9Op3OpTt3b9myBdOmTcOtt94KwNaT4xerhduJ9nPGlaR2zs2hStsE4v6dImHwQLInIiL/lpKSgh07diArKwtFRUWN9qr06NED33zzDfbu3Yt9+/Zh8uTJAdkD01IMN14gGW3DUrvOBQHgfBsiIrKZO3cu1Go1+vTpg9jY2Ebn0CxatAhRUVEYOnQoUlNTMWrUKFx++eVtXFv/xWEpD5OEFSi33Qh0fa5t7HMQ7ydFREQALr74Ymzbts1p27Rp0+qVS0lJwbp165y2zZo1y+m9fZjK3qNjtVrrDcGVlJS0rsJ+ij03HmawnIMkZAiVFvtL9FCrJAzswvk2REREbYXhxsOCzLbVUZWGeAiocGlSOEL17CAjIiJqKww3HhZktl3XJl+yDUUN7sYhKSIiorbEcONhQRZbuMk02YaiBqVwMjEREVFbYrjxMPuw1NHqcEgScCXDDRERUZtiuPGwIIst3OSKDuiVEI6IYG0zexAREZEnMdx4mH3OzRkRg8G85QIREVGbY7jxMPuwVK7owHBDRETkAww3nmSphN5aDgDIER14s0wiIiIfYLjxpNp7SpULA+Ji49AhVO/jChERkZKkpKRg8eLFjveSJOG7775rtHxWVhYkScLevXtb9b2eOk5b4dXlPMh+N/Bc0YHXtyEiIq/Lzc1FREQETCaTx445bdo0lJSUOIWm5ORk5ObmIiYmxmPf400MN55UG244JEVERG0hISEBsix7NNw0RK1WIyEhwavf4UkclvIgc3E2AFu4GcybZRIRUR3vvfcekpKSHDe6tBs3bhzuueceZGZmYty4cYiPj0doaCiuvPJKrF27tsljXjgstXPnTlx22WUwGAy44oor8OuvvzqVt1qtmDFjBrp27YqgoCD07NkTr732muPz+fPn46OPPsKKFSsgSRIkScKGDRsaHJbauHEjBg0aBL1ej8TERDz55JOoqalxfH7dddfhoYcewuOPP47o6GgkJCRg/vz57v/gWoA9Nx50Lvc4ggFUBsUjIcLg6+oQEQUOIQBL5fn3smx7b1YDKi//f7w2GJCkZovdeeed+Mtf/oL169fjhhtuAAAUFxdj1apVWLlyJcrLyzF27Fg899xz0Ov1+Pjjj5GamoojR46gc+fOzR6/vLwcN910E0aMGIFPPvkEJ06cwMMPP+xURpZldOrUCV9++SU6dOiArVu34r777kNiYiLGjx+PuXPn4tChQzAajfjwww8BANHR0cjJyXE6zpkzZzB27FhMmzYNH3/8MQ4fPoyZM2fCYDA4BZiPPvoIc+bMwY4dO7Bt2zZMmzYNV199NUaMGNFse1qD4caDqs/aem5CYlN8WxEiokBjqQSeT3K8VQGIbKvv/r8cQBfSbLGoqCiMGTMGn376qSPcfPXVV4iJicGwYcOgUqnQv39/R/lnn30W3377Lb7//nvMnj272eN/+umnkGUZH3zwAQwGAy655BKcPn0aDzzwgKOMVqvFggULHO+7du2Kbdu24YsvvsD48eMRGhqKoKAgmEymJoeh3n77bSQnJ+PNN9+EJEno1asXcnJy8MQTT+CZZ56BqjZQ9uvXD/PmzQMA9OjRA2+++SbS09O9Hm44LOVBuopcAEB8p24+rgkREbVHU6ZMwddff+2YI7Ns2TJMnDgRKpUK5eXlmDt3Lnr37o3IyEiEhobi0KFDOHXqlEvHPnz4MPr16weD4fzIwZAhQ+qVe+uttzBw4EDExsYiNDQU7733nsvfYXfo0CEMGTIEUp0eq6uvvhrl5eU4ffq0Y1u/fv2c9ktMTERBQYFb39US7LnxkIpqC6JrCgAJuOiinr6uDhFRYNEG23pQasmyDGNZGcLDwhy9CF79bhelpqZCCIEff/wRV155JX7++Wf861//AgDMnTsXaWlpeOWVV3DRRRchKCgId9xxB8xms8eq+vnnn2Pu3Ll49dVXMWTIEISFheHll1/Gjh07PPYddWm1zrcgkiSp3pwjb2C48ZD9GScxRLIl8fiOXX1cGyKiACNJzkNDsgxorbZt3g43bjAYDLjtttuwbNkyZGRkoGfPnrj88ssBAFu2bMG0adNw6623ArDNocnKynL52L169cInn3yC6upqR+/N9u3bncps2bIFQ4cOxYMPPujYlpmZ6VRGp9PBarU2+V29e/fG119/DSGEo/dmy5YtCAsLQ6dOnVyus7e0nzPu5y4ylAAAyqUwSDrXUzwREQWWKVOm4Mcff8SSJUswZcoUx/YePXrgm2++wd69e7Fv3z5MnjzZrV6OyZMnQ5IkzJw5E7///jtWrlyJV155xalMjx498Msvv2D16tU4evQonn76aezatcupTEpKCvbv348jR46gqKgIFoul3nc9+OCDyM7Oxl/+8hccPnwYK1aswLx58zBnzhzv95S5wPc1UIhYTTWEIQI1Bl7fhoiIGnf99dcjOjoaR44cweTJkx3bFy1ahKioKAwdOhSpqakYNWqUo1fHFaGhofjf//6HAwcO4LLLLsPf/vY3vPjii05l/vznP+O2227DhAkTMHjwYJw9e9apFwcAZs6ciZ49e+KKK65AbGwstmzZUu+7OnbsiJUrV2Lnzp3o378/7r//fsyYMQN///vf3fxpeAeHpTwl5Q+oeSwTP/+wAqN9XRciImq3VCpVvaXVgK3HZN26dU7bZs2a5fT+wmEqIYRtfpHRCAC46qqr6t0iQQjheK3X6/Hhhx86lnnbLVy40PE6NjYWa9asqVe/uscBgGuvvRY7d+6sV85uw4YN9bY1dasIT2LPjYfJKm3zhYiIiMhrGG6IiIhIURhuiIiISFEYboiIiEhRGG6IiIhIURhuiIjIb124gof8m6fOJ8MNERH5HbVaDQAevTUB+Z79fNrPb0vxOjdEROR3NBoNgoODUVhYCK1WW++quLIsw2w2o7q6ul1cMdeblNJWWZZRWFiI4OBgaDStiycMN0RE5HckSUJiYiJOnDiBkydP1vtcCIGqqioEBQU53blaiZTUVpVKhc6dO7e6HQw3RETkl3Q6HXr06NHg0JTFYsGmTZvwxz/+sd6dqZVGSW3V6XQe6X1iuCEiIr+lUqkcd8CuS61Wo6amBgaDwe//4DcnkNrqKv8dnCMiIiJqAMMNERERKQrDDRERESlKwM25sV8gyH57eE+yWCyorKyE0WhU/Lgn26pcgdRetlW5Aqm9gdJW+99tVy70F3DhpqysDACQnJzs45oQERGRu8rKyhAREdFkGUkE2LWrZVlGTk4OwsLCPH49AKPRiOTkZGRnZyM8PNyjx25v2FblCqT2sq3KFUjtDZS2CiFQVlaGpKSkZpeLB1zPjUqlQqdOnbz6HeHh4Yr+B1YX26pcgdRetlW5Aqm9gdDW5nps7DihmIiIiBSF4YaIiIgUheHGg/R6PebNmwe9Xu/rqngd26pcgdRetlW5Aqm9gdRWVwXchGIiIiJSNvbcEBERkaIw3BAREZGiMNwQERGRojDcEBERkaIw3LjprbfeQkpKCgwGAwYPHoydO3c2Wf7LL79Er169YDAY0LdvX6xcubKNatpyCxcuxJVXXomwsDDExcXhlltuwZEjR5rcZ+nSpZAkyelhMBjaqMatM3/+/Hp179WrV5P7+ON5BYCUlJR6bZUkCbNmzWqwvD+d102bNiE1NRVJSUmQJAnfffed0+dCCDzzzDNITExEUFAQhg8fjmPHjjV7XHd/59tKU+21WCx44okn0LdvX4SEhCApKQlTp05FTk5Ok8dsye9CW2ju3E6bNq1evUePHt3scdvjuW2urQ39/kqShJdffrnRY7bX8+pNDDduWL58OebMmYN58+Zhz5496N+/P0aNGoWCgoIGy2/duhWTJk3CjBkz8Ouvv+KWW27BLbfcgoMHD7Zxzd2zceNGzJo1C9u3b0daWhosFgtGjhyJioqKJvcLDw9Hbm6u43Hy5Mk2qnHrXXLJJU5137x5c6Nl/fW8AsCuXbuc2pmWlgYAuPPOOxvdx1/Oa0VFBfr374+33nqrwc9feuklvP7663j33XexY8cOhISEYNSoUaiurm70mO7+zrelptpbWVmJPXv24Omnn8aePXvwzTff4MiRI7j55pubPa47vwttpblzCwCjR492qvdnn33W5DHb67ltrq1125ibm4slS5ZAkiTcfvvtTR63PZ5XrxLkskGDBolZs2Y53lutVpGUlCQWLlzYYPnx48eLG2+80Wnb4MGDxZ///Gev1tPTCgoKBACxcePGRst8+OGHIiIiou0q5UHz5s0T/fv3d7m8Us6rEEI8/PDDonv37kKW5QY/99fzCkB8++23jveyLIuEhATx8ssvO7aVlJQIvV4vPvvss0aP4+7vvK9c2N6G7Ny5UwAQJ0+ebLSMu78LvtBQW++++24xbtw4t47jD+fWlfM6btw4cf311zdZxh/Oq6ex58ZFZrMZu3fvxvDhwx3bVCoVhg8fjm3btjW4z7Zt25zKA8CoUaMaLd9elZaWAgCio6ObLFdeXo4uXbogOTkZ48aNw2+//dYW1fOIY8eOISkpCd26dcOUKVNw6tSpRssq5byazWZ88sknuOeee5q8iaw/n1e7EydOIC8vz+m8RUREYPDgwY2et5b8zrdnpaWlkCQJkZGRTZZz53ehPdmwYQPi4uLQs2dPPPDAAzh79myjZZVybvPz8/Hjjz9ixowZzZb11/PaUgw3LioqKoLVakV8fLzT9vj4eOTl5TW4T15enlvl2yNZlvHII4/g6quvxqWXXtpouZ49e2LJkiVYsWIFPvnkE8iyjKFDh+L06dNtWNuWGTx4MJYuXYpVq1bhnXfewYkTJ3DNNdegrKyswfJKOK8A8N1336GkpATTpk1rtIw/n9e67OfGnfPWkt/59qq6uhpPPPEEJk2a1OSNFd39XWgvRo8ejY8//hjp6el48cUXsXHjRowZMwZWq7XB8ko5tx999BHCwsJw2223NVnOX89rawTcXcHJPbNmzcLBgwebHZ8dMmQIhgwZ4ng/dOhQ9O7dG//+97/x7LPPeruarTJmzBjH6379+mHw4MHo0qULvvjiC5f+j8hfffDBBxgzZgySkpIaLePP55VsLBYLxo8fDyEE3nnnnSbL+uvvwsSJEx2v+/bti379+qF79+7YsGEDbrjhBh/WzLuWLFmCKVOmNDvJ31/Pa2uw58ZFMTExUKvVyM/Pd9qen5+PhISEBvdJSEhwq3x7M3v2bPzwww9Yv349OnXq5Na+Wq0Wl112GTIyMrxUO++JjIzExRdf3Gjd/f28AsDJkyexdu1a3HvvvW7t56/n1X5u3DlvLfmdb2/swebkyZNIS0trstemIc39LrRX3bp1Q0xMTKP1VsK5/fnnn3HkyBG3f4cB/z2v7mC4cZFOp8PAgQORnp7u2CbLMtLT053+z7auIUOGOJUHgLS0tEbLtxdCCMyePRvffvst1q1bh65du7p9DKvVigMHDiAxMdELNfSu8vJyZGZmNlp3fz2vdX344YeIi4vDjTfe6NZ+/npeu3btioSEBKfzZjQasWPHjkbPW0t+59sTe7A5duwY1q5diw4dOrh9jOZ+F9qr06dP4+zZs43W29/PLWDreR04cCD69+/v9r7+el7d4usZzf7k888/F3q9XixdulT8/vvv4r777hORkZEiLy9PCCHEXXfdJZ588klH+S1btgiNRiNeeeUVcejQITFv3jyh1WrFgQMHfNUElzzwwAMiIiJCbNiwQeTm5joelZWVjjIXtnXBggVi9erVIjMzU+zevVtMnDhRGAwG8dtvv/miCW557LHHxIYNG8SJEyfEli1bxPDhw0VMTIwoKCgQQijnvNpZrVbRuXNn8cQTT9T7zJ/Pa1lZmfj111/Fr7/+KgCIRYsWiV9//dWxOuiFF14QkZGRYsWKFWL//v1i3LhxomvXrqKqqspxjOuvv1688cYbjvfN/c77UlPtNZvN4uabbxadOnUSe/fudfo9NplMjmNc2N7mfhd8pam2lpWViblz54pt27aJEydOiLVr14rLL79c9OjRQ1RXVzuO4S/ntrl/x0IIUVpaKoKDg8U777zT4DH85bx6E8ONm9544w3RuXNnodPpxKBBg8T27dsdn1177bXi7rvvdir/xRdfiIsvvljodDpxySWXiB9//LGNa+w+AA0+PvzwQ0eZC9v6yCOPOH4u8fHxYuzYsWLPnj1tX/kWmDBhgkhMTBQ6nU507NhRTJgwQWRkZDg+V8p5tVu9erUAII4cOVLvM38+r+vXr2/w3629PbIsi6efflrEx8cLvV4vbrjhhno/gy5duoh58+Y5bWvqd96XmmrviRMnGv09Xr9+veMYF7a3ud8FX2mqrZWVlWLkyJEiNjZWaLVa0aVLFzFz5sx6IcVfzm1z/46FEOLf//63CAoKEiUlJQ0ew1/OqzdJQgjh1a4hIiIiojbEOTdERESkKAw3REREpCgMN0RERKQoDDdERESkKAw3REREpCgMN0RERKQoDDdERESkKAw3RBRwUlJSsHjxYl9Xg4i8hOGGiLxq2rRpuOWWWwAA1113HR555JE2++6lS5ciMjKy3vZdu3bhvvvua7N6EFHb0vi6AkRE7jKbzdDpdC3ePzY21oO1IaL2hj03RNQmpk2bho0bN+K1116DJEmQJAlZWVkAgIMHD2LMmDEIDQ1FfHw87rrrLhQVFTn2ve666zB79mw88sgjiImJwahRowAAixYtQt++fRESEoLk5GQ8+OCDKC8vBwBs2LAB06dPR2lpqeP75s+fD6D+sNSpU6cwbtw4hIaGIjw8HOPHj0d+fr7j8/nz52PAgAH473//i5SUFERERGDixIkoKyvz7g+NiFqE4YaI2sRrr72GIUOGYObMmcjNzUVubi6Sk5NRUlKC66+/Hpdddhl++eUXrFq1Cvn5+Rg/frzT/h999BF0Oh22bNmCd999FwCgUqnw+uuv47fffsNHH32EdevW4fHHHwcADB06FIsXL0Z4eLjj++bOnVuvXrIsY9y4cSguLsbGjRuRlpaG48ePY8KECU7lMjMz8d133+GHH37ADz/8gI0bN+KFF17w0k+LiFqDw1JE1CYiIiKg0+kQHByMhIQEx/Y333wTl112GZ5//nnHtiVLliA5ORlHjx7FxRdfDADo0aMHXnrpJadj1p2/k5KSgn/+85+4//778fbbb0On0yEiIgKSJDl934XS09Nx4MABnDhxAsnJyQCAjz/+GJdccgl27dqFK6+8EoAtBC1duhRhYWEAgLvuugvp6el47rnnWveDISKPY88NEfnUvn37sH79eoSGhjoevXr1AmDrLbEbOHBgvX3Xrl2LG264AR07dkRYWBjuuusunD17FpWVlS5//6FDh5CcnOwINgDQp08fREZG4tChQ45tKSkpjmADAImJiSgoKHCrrUTUNthzQ0Q+VV5ejtTUVLz44ov1PktMTHS8DgkJcfosKysLN910Ex544AE899xziI6OxubNmzFjxgyYzWYEBwd7tJ5ardbpvSRJkGXZo99BRJ7BcENEbUan08FqtTptu/zyy/H1118jJSUFGo3r/0navXs3ZFnGq6++CpXK1gn9xRdfNPt9F+rduzeys7ORnZ3t6L35/fffUVJSgj59+rhcHyJqPzgsRURtJiUlBTt27EBWVhaKioogyzJmzZqF4uJiTJo0Cbt27UJmZiZWr16N6dOnNxlMLrroIlgsFrzxxhs4fvw4/vvf/zomGtf9vvLycqSnp6OoqKjB4arhw4ejb9++mDJlCvbs2YOdO3di6tSpuPbaa3HFFVd4/GdARN7HcENEbWbu3LlQq9Xo06cPYmNjcerUKSQlJWHLli2wWq0YOXIk+vbti0ceeQSRkZGOHpmG9O/fH4sWLcKLL76ISy+9FMuWLcPChQudygwdOhT3338/JkyYgNjY2HoTkgHb8NKKFSsQFRWFP/7xjxg+fDi6deuG5cuXe7z9RNQ2JCGE8HUliIiIiDyFPTdERESkKAw3REREpCgMN0RERKQoDDdERESkKAw3REREpCgMN0RERKQoDDdERESkKAw3REREpCgMN0RERKQoDDdERESkKAw3REREpCgMN0RERKQo/w/vO6maPRoOPAAAAABJRU5ErkJggg==",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# vizualize mnist\n",
    "\n",
    "for item in history.history:\n",
    "    print(\"Key:\",item)\n",
    "\n",
    "plt.plot(history.history['loss'],label=\"train\")\n",
    "plt.plot(history.history['val_loss'],label=\"validation\")\n",
    "plt.title('Model Loss')\n",
    "plt.yscale('log')\n",
    "plt.ylabel('Cross Entropy')\n",
    "plt.xlabel('Iteration')\n",
    "plt.grid()\n",
    "plt.legend()\n",
    "plt.show()\n",
    "\n",
    "#print(\"history\",history.history)\n",
    "plt.plot(history.history['sparse_categorical_accuracy'],label=\"train\")\n",
    "plt.plot(history.history['val_sparse_categorical_accuracy'],label=\"validation\")\n",
    "plt.title('Model Accuracy')\n",
    "#plt.yscale('log')\n",
    "plt.ylabel('Acc')\n",
    "plt.xlabel('Iteration')\n",
    "plt.grid()\n",
    "plt.legend()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 2.\n",
    "\n",
    "Realizuoti ResNet tinklą ir palyginti su konvoliuciniu tinklu be liekamųjų ryšių."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 84,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/10\n",
      "235/235 [==============================] - 74s 306ms/step - loss: 0.7743 - accuracy: 0.7702 - val_loss: 0.4372 - val_accuracy: 0.8668\n",
      "Epoch 2/10\n",
      "235/235 [==============================] - 72s 308ms/step - loss: 0.3905 - accuracy: 0.8771 - val_loss: 0.3402 - val_accuracy: 0.8938\n",
      "Epoch 3/10\n",
      "235/235 [==============================] - 71s 302ms/step - loss: 0.3226 - accuracy: 0.8979 - val_loss: 0.2689 - val_accuracy: 0.9129\n",
      "Epoch 4/10\n",
      "235/235 [==============================] - 74s 317ms/step - loss: 0.2788 - accuracy: 0.9112 - val_loss: 0.2699 - val_accuracy: 0.9119\n",
      "Epoch 5/10\n",
      "235/235 [==============================] - 70s 299ms/step - loss: 0.2666 - accuracy: 0.9151 - val_loss: 0.2245 - val_accuracy: 0.9277\n",
      "Epoch 6/10\n",
      "235/235 [==============================] - 70s 300ms/step - loss: 0.2364 - accuracy: 0.9244 - val_loss: 0.2107 - val_accuracy: 0.9305\n",
      "Epoch 7/10\n",
      "235/235 [==============================] - 70s 299ms/step - loss: 0.2248 - accuracy: 0.9284 - val_loss: 0.2309 - val_accuracy: 0.9260\n",
      "Epoch 8/10\n",
      "235/235 [==============================] - 75s 318ms/step - loss: 0.2125 - accuracy: 0.9319 - val_loss: 0.2109 - val_accuracy: 0.9304\n",
      "Epoch 9/10\n",
      "235/235 [==============================] - 73s 312ms/step - loss: 0.2070 - accuracy: 0.9337 - val_loss: 0.2505 - val_accuracy: 0.9124\n",
      "Epoch 10/10\n",
      "235/235 [==============================] - 76s 322ms/step - loss: 0.1955 - accuracy: 0.9379 - val_loss: 0.1950 - val_accuracy: 0.9356\n"
     ]
    }
   ],
   "source": [
    "from tensorflow.keras import datasets, layers, models, losses, Model\n",
    "\n",
    "mnist = tf.keras.datasets.mnist\n",
    "\n",
    "(train_images0, train_labels0),(test_images0,test_labels0) = datasets.mnist.load_data()\n",
    "train_images0 = tf.pad(train_images0, [[0, 0], [2,2], [2,2]])/255\n",
    "test_images0 = tf.pad(test_images0, [[0, 0], [2,2], [2,2]])/255\n",
    "train_images0 = tf.expand_dims(train_images0, axis=3, name=None)\n",
    "test_images0 = tf.expand_dims(test_images0, axis=3, name=None)\n",
    "train_images0 = tf.repeat(train_images0, 3, axis=3)\n",
    "test_images0 = tf.repeat(test_images0, 3, axis=3)\n",
    "\n",
    "test_images  = test_images0\n",
    "train_images = train_images0\n",
    "\n",
    "base_model = tf.keras.applications.ResNet50(weights = 'imagenet', include_top = False, input_shape = (32,32,3))\n",
    "\n",
    "for layer in base_model.layers:\n",
    "  layer.trainable = False\n",
    "\n",
    "x = layers.Flatten()(base_model.output)\n",
    "x = layers.Dense(1000, activation='relu')(x)\n",
    "predictions = layers.Dense(10, activation = 'softmax')(x)\n",
    "\n",
    "head_model = Model(inputs = base_model.input, outputs = predictions)\n",
    "head_model.compile(optimizer='adam', loss=losses.sparse_categorical_crossentropy, metrics=['accuracy'])\n",
    "\n",
    "history = head_model.fit(train_images, train_labels0, batch_size=256, epochs=10, validation_data=(test_images, test_labels0))\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 89,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Key: loss\n",
      "Key: accuracy\n",
      "Key: val_loss\n",
      "Key: val_accuracy\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAloAAAHHCAYAAABnS/bqAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjguMywgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy/H5lhTAAAACXBIWXMAAA9hAAAPYQGoP6dpAAB3H0lEQVR4nO3deVxU9f7H8dfMsO+ggCIo7oIL7qbmjppbZbmUVi5lt8JrZtavurfs3tu+mJZmZaa2mFYulZmluJu7Ypq7Yu4goiKgssz8/jhK4VKgMwzL+/l4zKPhcOaczzApb7/nez5fk81msyEiIiIidmd2dgEiIiIipZWCloiIiIiDKGiJiIiIOIiCloiIiIiDKGiJiIiIOIiCloiIiIiDKGiJiIiIOIiCloiIiIiDKGiJiIiIOIiClojI3zCZTLz44ouFft3BgwcxmUxMmzbN7jWJSMmgoCUiJcK0adMwmUyYTCZWrVp11fdtNhsRERGYTCZ69uzphApv3LJlyzCZTHzzzTfOLkVE7ExBS0RKFA8PD2bMmHHV9uXLl3PkyBHc3d2dUJWIyLUpaIlIidK9e3e+/vprcnJy8m2fMWMGTZo0oUKFCk6qTETkagpaIlKi3HvvvZw6dYpFixblbcvKyuKbb75hwIAB13xNRkYGTz75JBEREbi7u1O7dm3eeustbDZbvv0uXrzIE088QXBwML6+vtx+++0cOXLkmsc8evQoQ4cOJTQ0FHd3d+rWrcsnn3xivzd6DQcOHKBv374EBQXh5eXFLbfcwg8//HDVfu+99x5169bFy8uLwMBAmjZtmm8U8Ny5c4wcOZLIyEjc3d0JCQmhc+fObN682aH1i5RFCloiUqJERkbSsmVLvvzyy7xtP/74I2fPnuWee+65an+bzcbtt9/OO++8w2233cbYsWOpXbs2Tz31FKNGjcq370MPPcS4cePo0qULr732Gq6urvTo0eOqYyYlJXHLLbewePFihg8fzvjx46lRowYPPvgg48aNs/t7vnzOVq1a8dNPP/HYY4/x8ssvc+HCBW6//Xbmzp2bt9/kyZMZMWIE0dHRjBs3jv/85z80bNiQdevW5e3zyCOPMGnSJO6++27ef/99Ro8ejaenJzt37nRI7SJlmk1EpASYOnWqDbBt2LDBNmHCBJuvr68tMzPTZrPZbH379rV16NDBZrPZbFWqVLH16NEj73Xz5s2zAbaXXnop3/H69OljM5lMtn379tlsNpstISHBBtgee+yxfPsNGDDABtjGjBmTt+3BBx+0VaxY0ZaSkpJv33vuucfm7++fV1diYqINsE2dOvUv39vSpUttgO3rr7++7j4jR460AbaVK1fmbTt37pytatWqtsjISFtubq7NZrPZ7rjjDlvdunX/8nz+/v62uLi4v9xHROxDI1oiUuL069eP8+fPM3/+fM6dO8f8+fOve9lwwYIFWCwWRowYkW/7k08+ic1m48cff8zbD7hqv5EjR+b72mazMXv2bHr16oXNZiMlJSXv0bVrV86ePeuQS3ALFiygefPm3HrrrXnbfHx8ePjhhzl48CA7duwAICAggCNHjrBhw4brHisgIIB169Zx7Ngxu9cpIvkpaIlIiRMcHExsbCwzZsxgzpw55Obm0qdPn2vu+/vvvxMWFoavr2++7VFRUXnfv/xfs9lM9erV8+1Xu3btfF+fPHmSM2fO8NFHHxEcHJzvMWTIEACSk5Pt8j6vfB9X1nKt9/F///d/+Pj40Lx5c2rWrElcXByrV6/O95o33niD7du3ExERQfPmzXnxxRc5cOCA3WsWEXBxdgEiIjdiwIABDBs2jBMnTtCtWzcCAgKK5LxWqxWA++67j0GDBl1znwYNGhRJLdcSFRXF7t27mT9/PgsXLmT27Nm8//77vPDCC/znP/8BjBHBNm3aMHfuXH7++WfefPNNXn/9debMmUO3bt2cVrtIaaQRLREpkXr37o3ZbGbt2rXXvWwIUKVKFY4dO8a5c+fybd+1a1fe9y//12q1sn///nz77d69O9/Xl+9IzM3NJTY29pqPkJAQe7zFq97HlbVc630AeHt7079/f6ZOncqhQ4fo0aNH3uT5yypWrMhjjz3GvHnzSExMpFy5crz88st2r1ukrFPQEpESycfHh0mTJvHiiy/Sq1ev6+7XvXt3cnNzmTBhQr7t77zzDiaTKW8E5/J/33333Xz7XXkXocVi4e6772b27Nls3779qvOdPHnyRt7O3+revTvr169nzZo1edsyMjL46KOPiIyMJDo6GoBTp07le52bmxvR0dHYbDays7PJzc3l7Nmz+fYJCQkhLCyMixcvOqR2kbJMlw5FpMS63qW7P+vVqxcdOnTgX//6FwcPHiQmJoaff/6Zb7/9lpEjR+bNyWrYsCH33nsv77//PmfPnqVVq1bEx8ezb9++q4752muvsXTpUlq0aMGwYcOIjo4mNTWVzZs3s3jxYlJTU2/o/cyePTtvhOrK9/nMM8/w5Zdf0q1bN0aMGEFQUBDTp08nMTGR2bNnYzYb/27u0qULFSpUoHXr1oSGhrJz504mTJhAjx498PX15cyZM4SHh9OnTx9iYmLw8fFh8eLFbNiwgbfffvuG6haRv+Dcmx5FRArmz+0d/sqV7R1sNqMNwhNPPGELCwuzubq62mrWrGl78803bVarNd9+58+ft40YMcJWrlw5m7e3t61Xr162w4cPX9XewWaz2ZKSkmxxcXG2iIgIm6urq61ChQq2Tp062T766KO8fQrb3uF6j8stHfbv32/r06ePLSAgwObh4WFr3ry5bf78+fmO9eGHH9ratm1rK1eunM3d3d1WvXp121NPPWU7e/aszWaz2S5evGh76qmnbDExMTZfX1+bt7e3LSYmxvb+++//ZY0icmNMNtsVrZFFRERExC40R0tERETEQRS0RERERBxEQUtERETEQRS0RERERBxEQUtERETEQRS0RERERBxEDUudyGq1cuzYMXx9fTGZTM4uR0RERArAZrNx7tw5wsLC8poFX4+ClhMdO3aMiIgIZ5chIiIiN+Dw4cOEh4f/5T4KWk7k6+sLGB+Un5+fXY+dnZ3Nzz//TJcuXXB1dbXrsaXw9HkUL/o8ihd9HsWLPo+/l5aWRkRERN7v8b+ioOVEly8X+vn5OSRoeXl54efnpz8oxYA+j+JFn0fxos+jeNHnUXAFmfajyfAiIiIiDqKgJSIiIuIgCloiIiIiDqI5WiIiInaQm5tLdna2s8u4adnZ2bi4uHDhwgVyc3OdXY5TuLq6YrFY7HIsBS0REZGbYLPZOHHiBGfOnHF2KXZhs9moUKEChw8fLtM9HgMCAqhQocJN/wwUtERERG7C5ZAVEhKCl5dXiQ8nVquV9PR0fHx8/rYZZ2lks9nIzMwkOTkZgIoVK97U8RS0nGDixIlMnDixzA7JioiUFrm5uXkhq1y5cs4uxy6sVitZWVl4eHiUyaAF4OnpCUBycjIhISE3dRmxbP4EnSwuLo4dO3awYcMGZ5ciIiI34fKcLC8vLydXIvZ2+TO92Xl3CloiIiI3qaRfLpSr2eszVdASERERcRAFLREREbkpkZGRjBs3ztllFEuaDC8iIlIGtW/fnoYNG9olIG3YsAFvb++bL6oUUtAqpY6cPk/KBWdXISIiJZXNZiMnJwcXl7+PCsHBwUVQUcmkS4el0Hvxe+kwdiWLj+rjFRGRqw0ePJjly5czfvx4TCYTJpOJadOmYTKZ+PHHH2nfvj2enp6sWrWK/fv3c8cddxAaGoqPjw/NmjVj8eLF+Y535aVDk8nExx9/TO/evfHy8qJmzZp89913Rfwuiwf9Ji6FYiICAPjttAmr1ebcYkREyhibzUZmVo5THjZbwf7OHz9+PC1btmTYsGEcP36c48ePExERAcBzzz3HmDFj+O2332jQoAHp6el0796d+Ph4tmzZwm233UavXr04dOjQX57jP//5D/369ePXX3+le/fuDBw4kNTU1Jv++ZY0unRYCrWoFoS3m4W0rFy2H0ujSdXyzi5JRKTMOJ+dS/QLPznl3Dv+2xUvt7//1e7v74+bmxteXl5UqFABgF27dgHw4osv0qFDB/z8/DCbzQQFBRETE5P32v/973/MnTuX7777juHDh1/3HIMHD+bee+8F4JVXXuHdd99l/fr13HbbbTfzFkscjWiVQu4uFm6tYXQoXrL7pJOrERGRkqRp06b5vk5PT2f06NFERUUREBCAj48PO3fu/NsRrQYNGuQ99/b2xs/PL29Zm7JEI1qlVMc6wfy0I5klu07y1G1Rzi5HRKTM8HS1sOO/XZ127pt15d2Do0ePZtGiRbz11lvUqFEDT09P+vTpQ1ZW1l8ex9XVNd/XJpMJq9V60/WVNApapVS7WsGYsLHzxDmOnTlPWICns0sSESkTTCZTgS7fOZubm1uB1txdvXo1gwcPpnfv3oAxwnXw4EEHV1d66NJhKVXO241IX+N5/K6yN1QrIiJ/LTIyknXr1nHw4EFSUlKuO9pUs2ZN5syZQ0JCAlu3bmXAgAFlcmTqRilolWJ1A40/CEt2Jjm5EhERKW5Gjx6NxWIhOjqa4ODg6865Gjt2LIGBgbRq1YpevXrRtWtXGjduXMTVllzFf2xTbljdQBvzD8Hq/afIzMopEUPZIiJSNGrVqsWaNWvybRs8eDBWq5W0tLS8bZGRkSxZsiTffnFxcfm+vvJS4rXaTJw5c+bmCi6hNKJVilX0hPAAD7JyrKzam+LsckRERMocBa1SzGSCDnVCAIjfqXlaIiIiRU1Bq5TrWNtYfyp+V7K6xIuIiBQxBa1SrnlkID7uLqSkX+TXo2edXY6IiEiZoqBVyrm5mGlby1iCJ153H4qIiBQpBa0yoFOdUAAWa56WiIhIkVLQKgM61AnBbIKdx9M4eua8s8sREREpMxS0yoAgbzcaVw4EYIm6xIuIiBQZBa0yomPU5TYPmqclIiJSVBS0yojYKGOe1i+XusSLiIjcjMjISMaNG5f3tclkYt68edfd/+DBg5hMJhISEm7qvPY6TlHRmixlRM0QHyKCPDmcep6Ve1PoWreCs0sSEZFS5Pjx4wQGBtr1mIMHD+bMmTP5AlxERATHjx+nfPnydj2Xo2hEq4wwmUx5dx/q8qGIiNhbhQoVcHd3d/h5LBYLFSpUwMWlZIwVKWiVIZcvHy7ZdVJd4kVEyrCPPvqIsLAwrFZrvu133HEHDz74IImJidx5552Ehobi4+NDs2bNWLx48V8e88pLh+vXr6dRo0Z4eHjQtGlTtmzZkm//3NxcHnzwQapWrYqnpye1a9dm/Pjxed9/8cUXmT59Ot9++y0mkwmTycSyZcuueelw+fLlNG/eHHd3dypWrMgzzzxDTs4f02Tat2/PiBEjePrppwkKCqJChQq8+OKLhf/B3QAFrTKkedUgfC91id965IyzyxERKZ1sNsjKcM7DVrB/RPft25dTp06xdOnSvG2pqaksXLiQAQMGkJ6eTrdu3YiPj2fLli3cdttt9OrVi0OHDhXo+Onp6fTs2ZPo6Gg2bdrEiy++yOjRo/PtY7VaCQ8P5+uvv2bHjh288MILPPfcc3z11VcAjB49mn79+nHbbbdx/Phxjh8/TqtWra4619GjR+nevTvNmjVj69atTJo0iSlTpvDSSy/l22/69Ol4e3uzbt063njjDf773/+yaNGiAr2fm1Eyxt3ELowu8cH8sO048TuTaVTZvtfSRUQEyM6EV8Kcc+7njoGb99/uFhgYSLdu3ZgxYwadOnUC4JtvvqF8+fJ06NCB9PR0WrdujdlsjMf873//Y+7cuXz33XcMHz78b48/Y8YMrFYrU6ZMwcPDg7p163LkyBEeffTRvH1cXV35z3/+k/d11apVWbNmDV999RX9+vXDx8cHT09PLl68SIUK159X/P777xMREcGECRMwmUzUqVOHY8eO8X//93+88MILee+hQYMGjBkzBoCaNWsyYcIE4uPj6dy589++n5uhEa0yptOlNg+LNU9LRKRMGzhwILNnz+bixYsAfPHFF9xzzz2YzWbS09N56qmniIqKIiAgAB8fH3bu3FngEa2dO3fSoEEDPDw88ra1bNnyqv0mTpxIkyZNCA4OxsfHh48++qjA5/jzuVq2bInJZMrb1rp1a9LT0zly5EjetgYNGuR7XcWKFUlOdnxvSY1olTHtaxtd4nedOMfRM+epFODp7JJEREoXVy9jZMlZ5y6gXr16YbPZ+OGHH2jWrBkrV67knXfeAeD5559nxYoVvPXWW9SoUQNPT0/69OlDVlaW3UqdOXMmo0eP5u2336Zly5b4+vry5ptvsm7dOrud489cXV3zfW0yma6ao+YIClpOMHHiRCZOnEhubm6Rn/tyl/iNv59myc4k7m8ZWeQ1iIiUaiZTgS7fOZuHhwd33XUXX3zxBfv27aN27do0btwYq9XKunXrGDRoEL179waMOVcHDx4s8LGjoqL47LPPuHDhQt6o1tq1a/Pts3r1alq1asVjjz2Wt23//v359nFzc/vb35VRUVHMnj0bm82WN6q1evVqfH19CQ8PL3DNjqJLh04QFxfHjh072LBhg1PO3ylKi0yLiIhx+fCHH37gk08+YeDAgXnbq1evzty5c0lISGDr1q0MGDCgUKM/AwYMwGQyMWzYMHbs2MGCBQt466238u1Ts2ZNNm7cyE8//cSePXt4/vnnr/q9GBkZya+//sru3btJSUkhOzv7qnM99thjHD58mH/+85/s2rWLb7/9ljFjxjBq1Ki8+VnO5PwKpMjFXpqntWb/KTIuqku8iEhZ1bFjR4KCgti9ezcDBgzI2/7yyy8TGBhIq1at6NWrF127dqVx48YFPq6Pjw/ff/8927Zto1GjRvzrX//i9ddfz7fPP/7xD+666y769+9PixYtOHXqVL7RLYBhw4ZRu3ZtmjZtSnBwMKtXr77qXJUqVWLBggWsX7+emJgYHnnkER588EH+/e9/F/Kn4Rgmm62A94KK3aWlpeHv78/Zs2fx8/Oz67Gzs7NZsGAB3bt3v+q6tM1mo92byziUmskH9zXhtnrqEu9of/V5SNHT51G8lOTP48KFCyQmJlK1atV8E79LMqvVSlpaGn5+fsViRMhZ/uqzLczv77L7EyzDTCZT3t2H6hIvIiLiOApaZdTlLvFLdyerS7yIiIiDKGiVUc0iL3eJzyJBXeJFREQcQkGrjHJzMdO2djAAS3T3oYiIiEMoaJVhseoSLyJiF7qvrPSx12eqoFWGta/1R5f4I6cznV2OiEiJc/kuycxM/R1a2lz+TG/2Tlh1hi/DAr3daFIlkA0HT7NkVzIPqEu8iEihWCwWAgIC8tbM8/LyyrfmXklktVrJysriwoULZbK9g81mIzMzk+TkZAICArBYLDd1PAWtMq5TVCgbDp5m8U4FLRGRG1GhgtGLsCgWKC4KNpuN8+fP4+npWeJD480ICAjI+2xvhoJWGRcbFcJrP+5i7f5TpF/Mwcdd/0uIiBSGyWSiYsWKhISEXHOJmJImOzubFStW0LZt2xLXQNZeXF1db3ok6zL9Vi3jqgf7UKWcF7+fymTV3pPcVq+is0sSESmRLBaL3X45O5PFYiEnJwcPD48yG7TsqexdfJV8TCYTnepokWkRERFHUNCSvDYPS3clk6su8SIiInajoCU0qxqEr4cLpzKy2Kou8SIiInajoCW4Wsy0q2V0idci0yIiIvajoCUAdLp0+TBe87RERETsRkFLAHWJFxERcQQFLQGMLvFNqwQBGtUSERGxFwUtydNJi0yLiIjYlYKW5OkUZfTTWncglfSLOU6uRkREpORT0JI81YO9iSznRVaulZV7Tjq7HBERkRJPQUvymEymvFEtdYkXERG5eQpaks/leVrLdqtLvIiIyM1S0JJ8mkX+0SU+4fAZZ5cjIiJSoiloST7qEi8iImI/ClpyldhL87TUT0tEROTmKGjJVdrXDsZiNrE76RyHU9UlXkRE5EYpaMlVArzcaFIlENDlQxERkZuhoGVHvXv3JjAwkD59+ji7lJsWe3mR6V26fCgiInKjFLTs6PHHH+fTTz91dhl2cbmf1toDpzh3IdvJ1YiIiJRMClp21L59e3x9fZ1dhl1UD/ahanlvsnNtrNyb4uxyRERESqRiEbSOHj3KfffdR7ly5fD09KR+/fps3LjRbsdfsWIFvXr1IiwsDJPJxLx5866538SJE4mMjMTDw4MWLVqwfv16u9VQEnWqc+nyoe4+FBERuSFOD1qnT5+mdevWuLq68uOPP7Jjxw7efvttAgMDr7n/6tWryc6++lLWjh07SEq69sTtjIwMYmJimDhx4nXrmDVrFqNGjWLMmDFs3ryZmJgYunbtSnLyHyGjYcOG1KtX76rHsWPHCvmuS4aOl+ZpLVWXeBERkRvi4uwCXn/9dSIiIpg6dWretqpVq15zX6vVSlxcHDVr1mTmzJlYLBYAdu/eTceOHRk1ahRPP/30Va/r1q0b3bp1+8s6xo4dy7BhwxgyZAgAH3zwAT/88AOffPIJzzzzDAAJCQk38hZLrMtd4lMzskg4fJomVYKcXZKIiEiJ4vQRre+++46mTZvSt29fQkJCaNSoEZMnT77mvmazmQULFrBlyxYeeOABrFYr+/fvp2PHjtx5553XDFkFkZWVxaZNm4iNjc13rtjYWNasWXNDx/wrEydOJDo6mmbNmtn92PbkajHTvrYxqqVFpkVERArP6UHrwIEDTJo0iZo1a/LTTz/x6KOPMmLECKZPn37N/cPCwliyZAmrVq1iwIABdOzYkdjYWCZNmnTDNaSkpJCbm0toaGi+7aGhoZw4caLAx4mNjaVv374sWLCA8PDw64a0uLg4duzYwYYNG2645qKS1+ZB/bREREQKzemXDq1WK02bNuWVV14BoFGjRmzfvp0PPviAQYMGXfM1lStX5rPPPqNdu3ZUq1aNKVOmYDKZirLsa1q8eLGzS7C79rVCsJhN7ElK53BqJhFBXs4uSUREpMRw+ohWxYoViY6OzrctKiqKQ4cOXfc1SUlJPPzww/Tq1YvMzEyeeOKJm6qhfPnyWCyWqybTJyUlUaFChZs6dknn7+VK00td4hdrVEtERKRQnB60Wrduze7du/Nt27NnD1WqVLnm/ikpKXTq1ImoqCjmzJlDfHw8s2bNYvTo0Tdcg5ubG02aNCE+Pj5vm9VqJT4+npYtW97wcUsLLTItIiJyY5wetJ544gnWrl3LK6+8wr59+5gxYwYfffQRcXFxV+1rtVrp1q0bVapUYdasWbi4uBAdHc2iRYuYOnUq77zzzjXPkZ6eTkJCQt5dg4mJiSQkJOQbNRs1ahSTJ09m+vTp7Ny5k0cffZSMjIy8uxDLsk6X5mmtS1SXeBERkcJw+hytZs2aMXfuXJ599ln++9//UrVqVcaNG8fAgQOv2tdsNvPKK6/Qpk0b3Nzc8rbHxMSwePFigoODr3mOjRs30qFDh7yvR40aBcCgQYOYNm0aAP379+fkyZO88MILnDhxgoYNG7Jw4cKrJsiXRdWCfahW3psDKRms3JtC9/oVnV2SiIhIieD0oAXQs2dPevbsWaB9O3fufM3tjRo1uu5r2rdvj8329w03hw8fzvDhwwtUR1nTsU4IB1YlsnhnkoKWiIhIATn90qGUDJcXmV62+6S6xIuIiBSQgpYUSNPIQPwudYnfcui0s8sREREpERS0pEDUJV5ERKTwFLSkwDqpS7yIiEihKGhJgV3uEr83OZ1DpzKdXY6IiEixp6AlBebv5UqzSHWJFxERKSgFLSmUvC7xuxS0RERE/o6ClhTK5TYP6w6kqku8iIjI31DQkkKpWt6bauW9ybHaWLEnxdnliIiIFGsKWlJouvtQRESkYBS0pNAuXz5cujtZXeJFRET+goKWFFrTKoH4e7pyOjObzeoSLyIicl0KWlJoLhYz7WsHA2rzICIi8lcUtOSGXL58GK/leERERK5LQUtuSLtawbiYTexLTuf3UxnOLkdERKRYUtCSG+Lv6UqzyCBAi0yLiIhcj4KW3LDLbR6WqEu8iIjINSloyQ37c5f4NHWJFxERuYqCltywquW9qRZ8uUv8SWeXIyIiUuwoaMlNidXdhyIiIteloCU3pVMdY57W0t3J5ORanVyNiIhI8aKgJTelyaUu8Wcys9l86IyzyxERESlWFLTkprhYzHS41CVei0yLiIjkp6AlN+3y3YdajkdERCQ/BS25ae1qG13i95/M4GCKusSLiIhcpqAlN83Pw5XmVY0u8fG7dPehiIjIZQpaYhcdL919qHlaIiIif1DQEru43E9rfaK6xIuIiFymoCV2EVnem+qXusQv360u8SIiIqCgJXb0R5d4XT4UEREBBS2xo8ttHpbuPqku8SIiIihoiR01rhxAgJcrZ89ns+n3084uR0RExOkUtMRujC7xl+4+VJsHERERBS2xr05RavMgIiJymYKW2FXbWuoSLyIicpmCltjVn7vEa+1DEREp6xS0nGDixIlER0fTrFkzZ5fiEJ3y2jxonpaIiJRtClpOEBcXx44dO9iwYYOzS3GI2EvztDYcTOXseXWJFxGRsktBS+yuSjlvaoT4GF3i96hLvIiIlF0KWuIQuvtQREREQUsc5PJyPMvUJV5ERMowBS1xiMaVAwm81CV+o7rEi4hIGaWgJQ5hMZvyusQvUZd4EREpoxS0xGE6XpqnpX5aIiJSViloicNc7hJ/4GQGieoSLyIiZZCCljiMn4crLaoZXeJ196GIiJRFClriUJ3qGHcf6vKhiIiURQpa4lCX2zxsOHias5nqEi8iImWLgpY4VOVyXtQM8SHXamPZHt19KCIiZYuCljicFpkWEZGySkFLHO7yItPLdieTrS7xIiJShhQ6aEVGRvLf//6XQ4cOOaIeKYUaXeoSn3Yhh03qEi8iImVIoYPWyJEjmTNnDtWqVaNz587MnDmTixcvOqI2KSX+3CVebR5ERKQsuaGglZCQwPr164mKiuKf//wnFStWZPjw4WzevNkRNUopoHlaIiJSFt3wHK3GjRvz7rvvcuzYMcaMGcPHH39Ms2bNaNiwIZ988gk2m82edUoJ17ZWeVwtJg6kZHDgZLqzyxERESkSNxy0srOz+eqrr7j99tt58sknadq0KR9//DF33303zz33HAMHDrRnnVLC+Xq40qJqOUCjWiIiUna4FPYFmzdvZurUqXz55ZeYzWYeeOAB3nnnHerUqZO3T+/evWnWrJldC5WSr1NUCKv2pbB4ZxLD2lZzdjkiIiIOV+gRrWbNmrF3714mTZrE0aNHeeutt/KFLICqVatyzz332K1IKR0ud4nf+Lu6xIuISNlQ6BGtAwcOUKVKlb/cx9vbm6lTp95wUVI6RQR5USvUhz1J6Szbk8wdDSs5uyQRERGHKnTQuhyyNm7cyM6dOwGIioqiadOm9q1MSqVOUaHsSUpn8U4FLRERKf0KHbSOHDnCvffey+rVqwkICADgzJkztGrVipkzZxIeHm7vGqUUiY0KYdKy/Sy/1CXe1aLFCUREpPQq9G+5hx56iOzsbHbu3Elqaiqpqans3LkTq9XKQw895IgapRRpGBFIkLcbaRdy2HhQXeJFRKR0K3TQWr58OZMmTaJ27dp522rXrs17773HihUr7FqclD4Ws4n2tYMBdYkXEZHSr9BBKyIiguzsq+8Yy83NJSwszC5FSel2+e7D+F3qpyUiIqVboYPWm2++yT//+U82btyYt23jxo08/vjjvPXWW3YtTkqnNjWNLvGJKRnsV5d4EREpxQo9GX7w4MFkZmbSokULXFyMl+fk5ODi4sLQoUMZOnRo3r6pqan2q1RKDV8PV26pVo6Ve1OI35lE9WAfZ5ckIiLiEIUOWuPGjXNAGVLWdKoTwsq9KSzemczDbas7uxwRERGHKHTQGjRokCPqEHvLPAXFeGHvTlGhvPj9Djb9fpozmVkEeLk5uyQRERG7K3TQAmPi+7x58/IaltatW5fbb78di8Vi1+LkBh1cjctX91M1qDvQw9nVXFNEkBe1Q33ZnXSOZbtPcmcjNS8VEZHSp9CT4fft20dUVBQPPPAAc+bMYc6cOdx3333UrVuX/fv3O6JGKaxjmzFlnqLekS8wJRbflhudokIA3X0oIiKlV6GD1ogRI6hevTqHDx9m8+bNbN68mUOHDlG1alVGjBjhiBqlsFoOx1q/H2asWOY+CKkHnF3RNV0OWssudYkXEREpbW6oYekbb7xBUFBQ3rZy5crx2muvsXz5crsWJzfIZCK3+1hOe1XDdP40fHkvXEhzdlVXudwl/tyFHDYc1B2qIiJS+hQ6aLm7u3Pu3Lmrtqenp+PmpgnNxYaLB+urPY7NpwKc3AVzHgZr8Ro1sphNdKh96fLhTl0+FBGR0qfQQatnz548/PDDrFu3DpvNhs1mY+3atTzyyCPcfvvtjqhRbtAF10By+34KFnfY8yMsfdnZJV0l9vI8rZ1J2IrxXZIiIiI3otBB691336V69eq0bNkSDw8PPDw8aN26NTVq1GD8+PGOqFFugi2sMdz+nvHFyrdg+2znFnSFNrWCcbOYOXgqk/0nM5xdjoiIiF0Vqr2DzWYjLS2NmTNncvTo0bz2DlFRUdSoUcMhBYodxPSH5N9g9XiYFwdB1SCskbOrAsDH3YUW1YLyusTXCFGXeBERKT0KNaJls9moUaMGR44coUaNGvTq1YtevXopZJUEncZAzS6Qcx5mDoRzSc6uKE/eItOapyUiIqVMoYKW2WymZs2anDp1ylH1iKOYLXD3x1CuJqQdha/uh5yLzq4K+KPNw8bfUzmdkeXkakREROyn0HO0XnvtNZ566im2b9/uiHrEkTz84d6Zxn8Pr4P5o4rFMj3hgV7UqeCL1QbL95x0djkiIiJ2U+ig9cADD7B+/XpiYmLw9PQkKCgo30OKufI1oM9UMJkh4XNY94GzKwKgYx1jVGvxzuJzSVNERORmFXqtw3feeQeTyeSIWqSo1OgEXV6Cn54zHsG1oXpHp5bUKSqU95ftZ/mek2TnWnG1FPrfACIiIsVOoYPW4MGDHVCGFLlbHoOk3yDhC/h6CAxbAuWqO62chhEBlPN241RGFhsSU2lVo7zTahEREbGXQg8bWCwWkpOvvjvs1KlTWCwWuxQlRcBkgp7vQHhzuHDm0jI9Z51WjsVsokPe5UPdfSgiIqVDoYPW9bp3X7x4UUvwlDQu7tD/c/ANg5TdMHsYWHOdVk5el/hd6hIvIiKlQ4EvHb777rsAmEwmPv74Y3x8/mgsmZuby4oVK6hTp479KxTH8g2Fe76Aqd1g70+w5H8Q+6JTSmlT0+gS//upTPafTKdGiK9T6hAREbGXAgetd955BzBGtD744IN8lwnd3NyIjIzkgw+Kxx1sztK7d2+WLVtGp06d+Oabb5xdTsFVagx3TITZD8KqdyCkLjToW+RleLu7cEv1cqzYc5LFO5MVtEREpMQr8KXDxMREEhMTadeuHVu3bs37OjExkd27d/PTTz/RokULR9Za7D3++ON8+umnzi7jxtTvA7eOMp5/NxyObnJKGX9eZFpERKSkK/QcraVLlxIYGOiIWkq89u3b4+tbgkdhOj4PtW6DnAuXluk5UfQlXJoQv+n30+oSLyIiJV6hg1Zubi5TpkxhwIABxMbG0rFjx3yPm/Haa69hMpkYOXLkTR3nSitWrKBXr16EhYVhMpmYN2/eNfebOHEikZGReHh40KJFC9avX2/XOoo9sxnumgzla8O540bYyr5QpCX8uUv8sj26+1BEREq2Qgetxx9/nMcff5zc3Fzq1atHTExMvseN2rBhAx9++CENGjT4y/1Wr15Ndnb2Vdt37NhBUtK1LzdlZGQQExPDxIkTr3vcWbNmMWrUKMaMGcPmzZuJiYmha9eu+VpZNGzYkHr16l31OHbsWAHfZQng4Qf3fgkeAXB0I8x/osiX6bm89qHaPIiISElX6IalM2fO5KuvvqJ79+52KyI9PZ2BAwcyefJkXnrppevuZ7VaiYuLo2bNmsycOTNvQv7u3bvp2LEjo0aN4umnn77qdd26daNbt25/WcPYsWMZNmwYQ4YMAeCDDz7ghx9+4JNPPuGZZ54BICEh4QbfYQlTrjr0mw6f3QVbZ0BoXWg1vMhO3ykqlIlL97Ni90mycqy4uahLvIiIlEyF/g3m5uZGjRo17FpEXFwcPXr0IDY29i/3M5vNLFiwgC1btvDAAw9gtVrZv38/HTt25M4777xmyCqIrKwsNm3alO/8ZrOZ2NhY1qxZc0PH/CsTJ04kOjqaZs2a2f3YdlOtPXR9xXi+6HnYt7jITt0wPIDyPm6cu5jDhoOpRXZeEREReyt00HryyScZP3683RpKzpw5k82bN/Pqq68WaP+wsDCWLFnCqlWrGDBgAB07diQ2NpZJkybdcA0pKSnk5uYSGhqab3toaCgnThR8QnhsbCx9+/ZlwYIFhIeHXzekxcXFsWPHDjZs2HDDNReJFv+ARveDzQpfD4WUfUVyWrPZRIfaWmRaRERKvkJfOly1ahVLly7lxx9/pG7duri6uub7/pw5cwp8rMOHD/P444+zaNEiPDw8Cvy6ypUr89lnn9GuXTuqVavGlClTisVC14sXF92oT5EwmaDH25CyFw6vhS/vgYcWg2eAw0/dKSqUrzcdIX5nMi/0jC4Wn6+IiEhhFXpEKyAggN69e9OuXTvKly+Pv79/vkdhbNq0ieTkZBo3boyLiwsuLi4sX76cd999FxcXF3Jzr70cTFJSEg8//DC9evUiMzOTJ554orBvI5/y5ctjsViumkyflJREhQoVburYJZ6LO/T/DPzC4dReo6lpESzT06ZmedwsZg6lZrIvOd3h5xMREXGEQo9oTZ061W4n79SpE9u2bcu3bciQIdSpU4f/+7//u+Yi1SkpKXTq1ImoqCi+/vpr9uzZQ/v27XF3d+ett966oTrc3Nxo0qQJ8fHx3HnnnYAx8T4+Pp7hw4tuEnix5RNiLNPzyW3GXK3FL0KX/zn0lN7uLrSsXo7le04SvyuZmqEluD+ZiIiUWQUe0fpzm4NrycnJKXTfKV9f36taJXh7e1OuXDnq1at31f5Wq5Vu3bpRpUoVZs2ahYuLC9HR0SxatIipU6fmLRN0pfT0dBISEvLuGkxMTCQhIYFDhw7l7TNq1CgmT57M9OnT2blzJ48++igZGRl5dyGWeWEN4c73jee/vAtbZzr8lJe7xE9bfZD9JzWqJSIiJU+Bg1bFihXzha369etz+PDhvK9PnTpFy5Yt7VvdFcxmM6+88gqzZ8/Gzc0tb3tMTAyLFy+mb99rr8+3ceNGGjVqRKNGjQAjVDVq1IgXXnghb5/+/fvz1ltv8cILL9CwYUMSEhJYuHDhVRPky7R6d0Hbp4zn342AI45dpueORpWoFuzNibQL9PtgDduPnnXo+UREROytwJcOr7zL8ODBg1c1DrXHnYjLli37y+937tz5mtsvh6hrad++fYFqGz58uC4V/p32z0HSDtj9A8wcAA8vA7+KDjmVn4crX/+jJYOmrmf70TTu/WgtHw9qSotq5RxyPhEREXuzaydI3RlWBpjNcNeHEBwF6SeMsJV93mGnK+fjzpfDbqFF1SDOXczhgU/Wa8FpEREpMdRyWwrP3ddYpsczEI5thu8fd+gyPb4erkwf2pzYqBAu5lh5+LNNzNty1GHnExERsZcCBy2TycS5c+dIS0vj7NmzmEwm0tPTSUtLy3tIGRJUFfpOB5MFfp1lTJB3IA9XC5Pua0LvRpXItdoYOSuB6b8cdOg5RUREblah5mjVqlUr39d/nhdls9l06bCsqdYOur0OC0bDojHG5cRaXRx2OleLmbf7xuDv6cq0Xw4y5rvfOHs+m392rKH/90REpFgqcNBaunSpI+uQkqrZQ5C0HTZNM5qZPhQPwbX+9mU3ymw2MaZXNAFeroxbvJexi/ZwOjOL53tEYzYrbImISPFS4KDVrl07R9YhJZXJBN3ehJN74NAvxjI9w+KN+VsOO6WJkbG18Pd05T/f72Dq6oOknc/h9bvr42LRtEMRESk+9FtJbp6LG/T7FPwjIHU/fPMg5OY4/LRDWldlbL8YLGYTszcf4dEvNnMh2/HLA4mIiBSUgpbYh08w3DMDXL1gfzwsHlMkp72rcTgf3tcENxczi3YkMWTqBtIvOj7kiYiIFISClthPxQbQ+wPj+ZoJsOWLIjltbHQo04c0x8fdhTUHTjFg8lpSM7KK5NwiIiJ/RUFL7Cv6Dmj3jPF8/kg4XLj1L29Uy+rl+HLYLQR5u/HrkbP0/eAXjp91XCNVERGRgrjpoJWWlsa8efPYuXOnPeqR0qDd/0GdnpCbBbPug7NF01y0frg/X/2jJRX9Pdh/MoM+k9ZwQItRi4iIExU6aPXr148JEyYAcP78eZo2bUq/fv1o0KABs2fPtnuBUgKZzdD7QwipC+lJDl+m589qhPjwzaOtqFbem6NnztPvwzX8dkyLUYuIiHMUOmitWLGCNm3aADB37lxsNhtnzpzh3Xff5aWXXrJ7gVJCufsYy/R4lYPjCfDtcIcu0/NnlQI8+eqRltQN8yMlPYt7PlzL+sTUIjm3iIjInxU6aJ09e5agoCAAFi5cyN13342Xlxc9evRg7969di9QSrDAKkbbB7MLbP8GVr1TZKcu7+POlw/fQvNLi1HfP2UdS3clF9n5RURE4AaCVkREBGvWrCEjI4OFCxfSpYux5Mrp06fx8PCwe4FSwkXeCt3eMJ7H/xd2LyyyU/t5uPLp0OZ0qmMsRj3s0418m6DFqEVEpOgUOmiNHDmSgQMHEh4eTlhYGO3btweMS4r169e3d31SGjR7EJo+CNhg9kOQvKvITu3hauGD+5twZ8Mwci4tRv3Z2t+L7PwiIlK2FTpoPfbYY6xZs4ZPPvmEVatWYTYbh6hWrZrmaMn1dXsdIttA1jljmZ7Mopsz5WoxM7ZfQwa1rILNBs/P286EJXuxFdGcMRERKbtuqL1D06ZN6d27Nz4+PuTm5pKQkECrVq1o3bq1veuT0sLiCn2nQ0BlOJ0I3wwpkmV6LjObTbx4e11GdKoJwFs/7+HlH3YqbImIiEPd0KXDKVOmAJCbm0u7du1o3LgxERERLFu2zN71SWniXQ7u+RJcveHAMvj530V6epPJxKjOtXi+ZzQAH69K5OlvfiUn11qkdYiISNlR6KD1zTffEBMTA8D3339PYmIiu3bt4oknnuBf//qX3QuUUqZCPbjrQ+P5ukmw+dMiL+HBW6vyVl9jMeqvNx0hboYWoxYREccodNBKSUmhQoUKACxYsIC+fftSq1Ythg4dyrZt2+xeoJRCUb2g/XPG8/mj4NDaIi+hT5NwJg1sjJuLmZ9+S2LoNC1GLSIi9lfooBUaGsqOHTvIzc1l4cKFdO7cGYDMzEwsFovdC5RSqu1TxrqI1mxjmZ4zh4u8hC51KzBtSDO83Sz8sv8UAyev5bQWoxYRETsqdNAaMmQI/fr1o169ephMJmJjYwFYt24dderUsXuBUkqZzXDnJAitDxknjWV6sjKLvIxW1cszY9gtBHq5svXIWfp9uIYTZy8UeR0iIlI6FTpovfjii3z88cc8/PDDrF69Gnd3dwAsFgvPPPOM3QuUUszNG+6dAV7l4cSv8G1ckS3T82cxEQF8/UhLKvh5sDc5nbsn/cLBlIwir0NEREqfG2rv0KdPH5544gnCw8Pztg0aNIg77rjDboVJGRFQGfp/BmZX+G0OrHzbKWXUCPHlm0dbUvXSYtR9PljDjmNpTqlFRERKjxsKWsuXL6dXr17UqFGDGjVqcPvtt7Ny5Up71yZlRZVW0OMt4/mS/8GuH5xSRnigF1/9oyXRFf1ISb9I/4/WsPGgFqMWEZEbV+ig9fnnnxMbG4uXlxcjRoxgxIgReHp60qlTJ2bMmOGIGqUsaDIYmj9sPJ/zMCTtcEoZwb7GYtTNIgM5dyGH+6asY+luLUYtIiI3ptBB6+WXX+aNN95g1qxZeUFr1qxZvPbaa/zvf/9zRI1SVnR9Baq2haz0Il+m58/8PV35dGgL2tcO5kK2lWHTN/Ld1mNOqUVEREq2QgetAwcO0KtXr6u233777SQmJtqlKCmjLi/TExgJZ36Hrx6A3GynlOLpZuGj+5tye4yxGPXjM7fwuRajFhGRQip00IqIiCA+Pv6q7YsXLyYiIsIuRUkZ5hVkLNPj5gMHV8JPzzmtFDcXM+P6N+S+Wypjs8G/521n4tJ9Wh9RREQKzKWwL3jyyScZMWJE3kLSAKtXr2batGmMHz/e7gVKGRQaDXdNNnprrf8IQqKh6RCnlGI2m/jfHfUI9HLjvSX7ePOn3Zw9n82z3epgMpmcUpOIiJQchQ5ajz76KBUqVODtt9/mq6++AiAqKopZs2apvYPYT53u0PFfsOQl+OFJCIiAGrFOKcVkMvFkl9r4e7ry0g87+WjFAc5kZvFK7/q4WG7oxl0RESkjChW0cnJyeOWVVxg6dCirVq1yVE0ihjaj4dR+2PolfDUIhvwIFRs4rZyH2lTDz9OVZ2b/ylcbj5B2Pofx9zbE3UVLT4mIyLUV6p/jLi4uvPHGG+TkaPFdKQImE/R69487Eb/o65Q1Ef+sX9MI3h/YBDeLmYW/neDBaRvJ0GLUIiJyHYW+7tGpUyeWL1/uiFpErubiBv0/N+ZppZ8wwtb5M04t6bZ6FZg6pBlebhZW7Uth4MfrOJOpxahFRORqhZ6j1a1bN5555hm2bdtGkyZN8Pb2zvf922+/3W7FiQDg4Q8Dv4aPY+HkTvjqfhg42whhTtK6hrEY9eCp60k4fIZ+H67hswdbEOrn4bSaRESk+Cl00HrssccAGDt27FXfM5lM5Obm3nxVIlfyD4cBX8HUbpC4Ar77J/T+wLi86CQNIwL46h8tuX/KOvYkGYtRf/FQC6qU8/77F4uISJlQ6EuHVqv1ug+FLHGoig2g33QwWeDXmbD0FWdXRK1QX755pBVVynlx5LSxGPXO41qMWkREDLo3XUqWGrHQa5zxfMUbsPlTp5YDEBHkxdePtKROBV9OnrtI/w/XsOl3LUYtIiKFCFpLliwhOjqatLSr/7V+9uxZ6taty4oVK+xanMg1NX4A2j5lPP9+JOxb7NRyAEJ8PZj1j5Y0qRJI2oUc7vt4Pcv3nHR2WSIi4mQFDlrjxo1j2LBh+Pn5XfU9f39//vGPf/DOO+/YtTiR6+rwL2hwD9hyjR5bx391dkX4e7ry2YPNaVcrmPPZuTw0fQPzf9Vi1CIiZVmBg9bWrVu57bbbrvv9Ll26sGnTJrsUJfK3TCa4/T2IbGP02JrRD84ecXZVeLm5MPmBpvRsUJHsXBv//HILX64/5OyyRETESQoctJKSknB1db3u911cXDh5UpdKpAhd7rEVHAXnjhs9ti6cdXZVuLmYGX9PIwa0MBajfnbONj5ckejsskRExAkKHLQqVarE9u3br/v9X3/9lYoVK9qlKJEC8wwwemz5VIDkHTDrfshxfvNQi9nEy3fW47H21QF4a9Fe5iSaycqxOrkyEREpSgUOWt27d+f555/nwoULV33v/PnzjBkzhp49e9q1OJECCYiAgV+Bmw8kLofvR4DN5uyqMJlMPH1bHZ7rXgeA5SfM9Ju8jn3J6U6uTEREikqBg9a///1vUlNTqVWrFm+88Qbffvst3377La+//jq1a9cmNTWVf/3rX46sVeT6KsZA30s9trZ+CctedXZFeR5uW533722It4uN346do+d7K/l87e/YikEYFBERxypw0AoNDeWXX36hXr16PPvss/Tu3ZvevXvz3HPPUa9ePVatWkVoaKgjaxX5azVjoeelO1+Xvw6bP3NuPX/SOTqE/4vJpXX1clzItvLvedsZ9ulGUtIvOrs0ERFxoEI1LK1SpQoLFiwgJSWFdevWsXbtWlJSUliwYAFVq1Z1VI0iBddk0J96bD1eLHpsXebvBp880Jjne0bjZjGzeGcyt41bydLdyc4uTUREHOSGOsMHBgbSrFkzmjdvTmBgoL1rErk5xbDH1mVms4kHb63Kt8NbUyvUh5T0iwyZuoEXv/uNC9lawkpEpLTREjxS+hTTHlt/FlXRj++G38rgVpEATPvlILdPWKV1EkVEShkFLSmdimmPrT/zcLXw4u11mTqkGeV93NmTlM4dE1bz8coDWK2aKC8iUhooaEnpVUx7bF2pQ+0QFo5sQ2xUCFm5Vl76YSeDpq4nKe3qVioiIlKyKGhJ6VZMe2xdqbyPO5MfaMpLd9bDw9XMyr0p3DZuBQu3n3B2aSIichMUtKT0qxgDfacVyx5bf2YymbjvlirM/2cb6ob5cTozm0c+38Qzs38l42KOs8sTEZEboKAlZUPNztBzrPG8mPXYulKNEB/mPtaaf7SrhskEMzccpud7q9h6+IyzSxMRkUJS0JKyo8lgaDPaeF7Memxdyc3FzLPdovjioRZU9PcgMSWDuyf9wsSl+8jVRHkRkRJDQcuOevfuTWBgIH369HF2KXI9Hf8NDfoXyx5b19KqenkWPt6WHvUrkmO18eZPu7n3o7UcOZ3p7NJERKQAFLTs6PHHH+fTTz91dhnyV0wmuH1Cse6xdSV/L1cmDGjEW31j8HazsP5gKt3Gr+TbhKPOLk1ERP6GgpYdtW/fHl9fX2eXIX8nr8dWnWLbY+tKJpOJPk3CWfB4GxpVDuDchRwen5nAE7MSSLuQ7ezyRETkOpwetCZNmkSDBg3w8/PDz8+Pli1b8uOPP9r1HCtWrKBXr16EhYVhMpmYN2/eNfebOHEikZGReHh40KJFC9avX2/XOqQY8QyAgd8U+x5bV6pSzpuv/9GSxzvVxGyCuVuO0m3cSjYcTHV2aSIicg1OD1rh4eG89tprbNq0iY0bN9KxY0fuuOMOfvvtt2vuv3r1arKzr/4X/I4dO0hKSrrmazIyMoiJiWHixInXrWPWrFmMGjWKMWPGsHnzZmJiYujatSvJyX8s+NuwYUPq1at31ePYsWOFfNdSLFzuseXqfanH1uPFssfWlVwsZp7oXIuvH2lJRJAnR8+cp/+Ha3j7591k51qdXZ6IiPyJ04NWr1696N69OzVr1qRWrVq8/PLL+Pj4sHbt2qv2tVqtxMXFMWDAAHJz/1iAd/fu3XTs2JHp06df8xzdunXjpZdeonfv3tetY+zYsQwbNowhQ4YQHR3NBx98gJeXF5988knePgkJCWzfvv2qR1hY2E38BMSpKsZAv+mXemzNgGWvObuiAmtSJYgFI9pwV+NKWG3w3pJ99PlgDQdTMpxdmoiIXOL0oPVnubm5zJw5k4yMDFq2bHnV981mMwsWLGDLli088MADWK1W9u/fT8eOHbnzzjt5+umnb+i8WVlZbNq0idjY2Hznio2NZc2aNTf8fq5n4sSJREdH06xZM7sfW25Avh5br8GWz51bTyH4ergytl9D3ru3EX4eLmw9fIbu767kqw2HsZWA0TkRkdKuWAStbdu24ePjg7u7O4888ghz584lOjr6mvuGhYWxZMkSVq1axYABA+jYsSOxsbFMmjTphs+fkpJCbm4uoaGh+baHhoZy4kTBl0CJjY2lb9++LFiwgPDw8OuGtLi4OHbs2MGGDRtuuGaxs6t6bMU7tZzC6hUTxsKRbWlRNYjMrFyenv0rj32xmdMZxX/emYhIaVYsglbt2rVJSEhg3bp1PProowwaNIgdO3Zcd//KlSvz2WefMWvWLFxcXJgyZQomk6kIK762xYsXc/LkSTIzMzly5Mg1R+WkGOv4b6jfD6w5Ro+tE9ucXVGhhAV4MmPYLfzfbXVwMZv4cfsJbhu/gtX7UpxdmohImVUsgpabmxs1atSgSZMmvPrqq8TExDB+/Pjr7p+UlMTDDz9Mr169yMzM5Iknnrip85cvXx6LxXLVZPqkpCQqVKhwU8eWEsRkgjsu99g6B1/0g7Mlq1eVxWzi0fbVmftYa6oFe5OUdpGBH6/jlQU7uZiT+/cHEBERuyoWQetKVquVixcvXvN7KSkpdOrUiaioKObMmUN8fDyzZs1i9OjRN3w+Nzc3mjRpQnz8H5eLrFYr8fHxGpUqa1zcof9nl3psHSsRPbaupX64P/P/eSsDW1QG4KMVB+g98Rf2JZ9zcmUiImWL04PWs88+y4oVKzh48CDbtm3j2WefZdmyZQwcOPCqfa1WK926daNKlSp5lw2jo6NZtGgRU6dO5Z133rnmOdLT00lISCAhIQGAxMREEhISOHToUN4+o0aNYvLkyUyfPp2dO3fy6KOPkpGRwZAhQxzyvqUY8wyEgV+DTygk/wZfPVAiemxdycvNhZd712fyA00J8nZjx/E0ery7is/WHNREeRGRIuLi7AKSk5N54IEHOH78OP7+/jRo0ICffvqJzp07X7Wv2WzmlVdeoU2bNri5ueVtj4mJYfHixQQHB1/zHBs3bqRDhw55X48aNQqAQYMGMW3aNAD69+/PyZMneeGFFzhx4gQNGzZk4cKFV02QlzIioDIM+AqmdocDy4wJ8ne+b1xeLGE6R4cSE9GG0V//yoo9J3n+299Yuvskr9/dgGBfd2eXJyJSqjk9aE2ZMqVQ+18rgAE0atTouq9p3759gf4FP3z4cIYPH16oeqQUC2to9Nia0d/osRVQGTo86+yqbkiIrwfTBjdj+pqDvPrjLpbsSqbb+BW82SeGDnVCnF2eiEip5fRLhyLFWs3O0ONt43kJ67F1JbPZxJDWVflueGvqVPAlJT2LIdM28MK327mQrYnyIiKOoKAl8neaDoE2TxrPS2CPrSvVqeDHvLjWDG1dFYBP1/xOz/dW8duxkjfpX0SkuFPQEimIjs9D/b4ltsfWlTxcLbzQK5pPhzYn2Nedfcnp9J74C5NXHMBq1UR5ERF7UdASKQiTCe6YWKJ7bF1L21rB/DSyLZ2jQ8nKtfLygp3c/8k6Tpy94OzSRERKBQUtkYIqJT22rhTk7cZH9zfh1bvq4+lqYfW+U9w2fgU/bjvu7NJEREo8BS2RwiglPbauZDKZuLd5ZX4YcSv1K/lzJjObR7/YzNPfbCXjYo6zyxMRKbEUtEQK63KPLVfvP3pslZIGoNWCfZj9aCsea18dkwm+2niEHu+uJOHwGWeXJiJSIiloidyIsIbQdxqYzEaPreWvO7siu3FzMfP0bXX4ctgthPl7cPBUJndP+oX34veSq4nyIiKFoqAlcqNqdYEeY43ny16FLV84tx47u6VaOX4c2ZZeMWHkWm28vWgP/T9cw96kc1rCR0SkgJzeGV6kRGs6BM4cglVj4fsR4FcRqnd0dlV24+/pyrv3NKRjnWCen/cbG38/Ted3VlDex42GEYE0qhxAo4gAGkQE4OOuv05ERK6kvxlFblbH5+HsYdj2Ncx6AIYuhAr1nF2V3ZhMJno3CqdplSD+PW87v+xPISU9i8U7k1i8M+nSPlArxJeGEQFG+KocSI0QHyzmkrc2pIiIPSloidwss9nosZV2HH5fZbR9eGgx+FdydmV2FRHkxfShzbmQncuO42lsOXSGLYdOk3D4DEdOn2d30jl2J51j1sbDAHi7WYi5FLwaRgTSMCJAi1iLSJmjoCViDy7ucM/nMKUrpOw2wtbQH8HD39mV2Z2Hq4XGlQNpXDkQMJbxST53gYRDZ0g4fIYth87w65EzZGTl8sv+U/yy/1Tea8MDPWlUOZBGEQE0rBxA3TA/3F0sTnonIiKOp6AlYi+egXDfN/Bx7B89tgZ+AxZXZ1fmcCG+HnSpW4EudSsAkGu1sTf5HFsOnSHh0Bm2HD7N3uR0jpw+z5HT5/l+6zEA3CxmosL8aHT5kmNEIBFBnphMuuQoIqWDgpaIPQVUhgGzYGqPP3ps3THR2VUVOYvZRJ0KftSp4Me9zSsDcO5CNr8eOZt3uXHLoTOcyshi6+EzbD18hmm/GK8t5+2WN9erYUQgDSL88fMo/WFVREonBS0RewtrZPTY+rI/JHxhhK/WTzq7Kqfz9XCldY3ytK5RHgCbzcaR0+fZfOi0MfJ1+Ay/HTvLqYws4nclE78rGTAm2tcM8bkUvow7HWuG+GqivYiUCApaIo5wucfW/JGw7FVMPmGAE+dr2WxgzQVrDlizITf70teXn+cYj2s+z4bcnD+2hzU0wuNNMplMRAR5ERHkxR0NjRsHLk+0Ny43niHh8GkOp55nT1I6e5LS+WrjEcCYaN8g3JjndXm+V4ivx03XJCJibwpaIo7ypx5blgVPUCv0dsxrEwHrFWHmUujJe55zKdhc6/mVQagQYcleXDygzydQp4f9jnlJ/on2hpPnLrL1sDHPy5hof5b0izmsOXCKNQf+mGhfKcDz0uVGY+SrbpgfHq6aaC8izqWgJeJIl3psmbZ9TdTxOXB8jrMrys9kMSbrm12MR95zV7C4XP38Qhqc2guz7oMeb0PToQ4vMdjXndjoUGKjQwFjov2+5PR8c732JJ/j6JnzHD1znvm/HgfA1WIiuqIfjSoH5s35qhzkpYn2IlKkFLREHOlSj61c/yoc376SsPDKmF3crgg1fxVwXMFs+dNzl2u85vLzS/vmPXe5dCzX65+vsKEjNwfmPw5bPof5T8C5E9D+2cIf5yZYzCZqV/CldgVf7vnTRPttR86y5VLwSjh8mpT0LLYeOcvWI2fzXht0aaJ9g0p+mNOLrGQRKcMUtEQczcUda7tn2JTRgNDu3TG7luA76CwucPsE8A2DFW8Yi2mfOw493jG+5yS+Hq60qlGeVldMtDeClzHy9dvRNFIzsliyK5klu5IBF3bZtvKvnnWpFODptNpFpHRT0BKRwjGZoOO/wLcCLBgNmz+F9JPGvC03L2dXB+SfaH97TBgAF3Ny2Xn8HFsOnWbt/hR+3pHEgu1JLNl9ksfa1+DhttU0p0tE7M7s7AJEpIRq9iD0+xQs7rDnR/j0dshMdXZV1+XuYqFhRABDWldlwr0NGd0gl6ZVAriQbWXsoj3Ejl3Owu3Hsdlszi5VREoRBS0RuXFRveCBb42lho5sgCld4PTvzq6qQMK9YcaDzXjv3kZU9PfgyOnzPPL5Zu6bso49SeecXZ6IlBIKWiJyc6q0hKE/gV8l447EKV3gxDZnV1UgJpOJXjFhxD/Zjn92rIGbi5nV+07RbfxK/vP9b5w9b8e2GCJSJiloicjNC4mCBxdBSDSkn4Cp3SFxhbOrKjAvNxee7FKbxU+0o2vdUHKtNqauPkiHt5bx5fpD5Fp1OVFEboyClojYh38lGLIAKreCi2nw+d2wvZj1Dfsblct58eH9TfnswebUCPEhNSOLZ+ds446Jq9h4sPjOPxOR4ktBS0TsxzMQ7p8LUbdDbhZ8MxTWfuDsqgqtTc1gfny8Dc/3jMbX3YXtR9Po88EaRs7cwomzF5xdnoiUIApaImJfrh7GotrNhgE2WPh/sOgFsFqdXVmhuFrMPHhrVZY+1Z57mkVgMsG8hGN0fHsZ7y/bx8WcXGeXKCIlgIKWiNif2QLd3zSWIAJYPR7mPWqsvVjClPdx57W7G/BtXGsaVw4gMyuXNxbupss7K1i8I0ntIETkLyloiYhjmEzQdjTc8b6xpuKvM2FGP7hYMlsnNAgP4JtHWjG2Xwwhvu78fiqThz7dyOCpG9h/Uuv5iMi1KWiJiGM1Ggj3zgRXL9i/BKb1hPRkZ1d1Q8xmE3c1DmfJ6PY80q46rhYTy/ecpOs7K3hlwU7OXSh5I3Yi4lgKWiLieLW6wKD54FUOjicYvbZO7Xd2VTfMx92FZ7rV4ecn2tGpTgg5VhsfrThAh7eW8/XGw1jVDkJELlHQEpGiEd4Ehv4MAZXhdKIRto5udnZVN6VqeW+mDG7G1MHNqFrem5T0izz1za/0nvQLCYfPOLs8ESkGFLREpOiUrwEPLoYKDSAzxbiMuG+xs6u6aR3qhPDTyLY8260O3m4Wth4+w50TV/PU11tJPqd2EFLyuOSed3YJpYaClogULd9QGPwDVG0H2Rkwoz9snensqm6am4uZf7SrztLR7bm7cTgAX286Qse3ljN5xQGyckpWewspo7Iyscx9iO6/PoJ5zXvOrqZUUNASkaLn4QcDv4H6fcGaA3P/AavGQSlolRDi58Hb/WKY81grGoT7k34xh5cX7OS28StYtrtk3gQgZUTacZjWHfOOeZiwYV7yX9i1wNlVlXgKWiLiHC5u0PsjaDnc+HrxGFj4TIlrbHo9jSsHMu+x1rzRpwHlfdw4cDKDwVM38ND0DRxMyXB2eSL5HdsCkzvAsS3YPIM47t8EEzaYMwySdji7uhJNQUtEnMdshq4vQ5eXja/XfQCzh0LORefWZSdms4l+TSNYMro9D91aFRezicU7k+nyzgpeX7iLjIs5zi5RBH6bB590g3PHIbgOOUN+ZkPVOKxVboWsdPjyHsjUWp83SkFLRJyv1XC4ewqYXeG3ucaC1BfOOrsqu/HzcOXfPaNZOLINbWqWJyvXyqRl++n49jLmbTmq7vLiHDYbLH8Tvh4EOeehRiw8+DMERmIzuZB71ycQUAXO/A5fPVAiV3YoDhS0RKR4qN8H7vsG3Hzh4EqY2t2YM1KK1Ajx5dOhzZn8QFMqB3mRlHaRkbMS6PvBGrYfLT3BUkqA7AvGZcGlLxlf3/IY3DsLPPz/2McryGg27OZj/Jlc+Kxzai3hFLREpPio1h6G/ADeIZC0HaZ0hpO7nV2VXZlMJjpHh/LzE215qmttPF0tbPz9NL0mrOLZOds4lV46LptKMXYuCab1gG1fg9kFer4Dt70KFper9w2NhrsmAybYMBk2flLk5ZZ0CloiUrxUjIGHFkFQdTh7GD7pCofXO7squ/NwtRDXoQZLRrfjjoZh2Gzw5fpDdHhrGVNXJ5KdWzpuCpBi5sQ2mNwRjm4EjwC4bw40HfrXr6nTHTr+23i+4Ck4uMrhZZYmCloiUvwERhpzRSo1gfOnYfrtpfY284r+noy/pxFfP9KSumF+pF3I4T/f76DHuytZvS/F2eVJabLrB5jSFdKOQLka8FA8VGtXsNe2eRLq3W20Y5l1P5w+6NBSSxMFLREpnrzLw6DvoWYXY6LurIGwaZqzq3KYZpFBfDf8Vl7pXZ9AL1f2JKUz8ON1PPLZJg6nZjq7PCnJbDajT93MgUaT4Krt4KHFxkoNBWUywe0ToGJDOJ8KXw6Ai+mOqrhUUdASkeLLzRvumQEN7wObFb5/HJa9Vioam16LxWxiQIvKLBvdgcGtIrGYTSz87QSxY5czdtEezmflOrtEKWlyLsK3cUafOmzQ9EG4bzZ4Bhb+WG5exp9H7xBI/s1oNFxK+t45koKWiBRvFle4YwK0GW18vexVmD8ScktvDyp/L1devL0uC0a0oVX1clzMsfJu/F46vb2M+b8eUzsIKZiMFPj0Dkj4Akxm6PYm9Bxr/Jm6Uf6V4J4vwOIGu+bD8tfsV28ppaAlIsWfyQSdnofubwEm4xLiV/dDVum+pFa7gi9fPNSCSQMbUynAk2NnLzB8xhbu+WgtO4+nObs8Kc6Sdhid3g+tAXd/Y8mrFg/b59gRzaHXeOP58teN3ndyXQpaIlJyNB8G/T4FizvsXgCf3VnqO1abTCa61a9I/JPteCK2Fu4uZtYlptLj3ZU8P287pzOynF2iFDd7foYpXeDMIQisatzFW6OTfc/RcMAfy2fNfRSOb7Xv8UsRBS0RKVmib4cH5hmNFQ+vg09ugzOHnV2Vw3m4Wng8tibxT7ajR/2KWG3w2drfaflaPAMmr2X84r2s2X+KC9max1Vm2WywZiJ82R+yzkGVW2HYEgiu7Zjzdf6v0U0+57wxOT5di6ZfyzW6k4mIFHNVWsHQn4ylelJ2G41N75sNoXWdXZnDhQd6MXFgY+7bf4r/fP8bu06c45f9p/hl/ykA3CxmGkYE0LxqEC2qBdG4ciDe7vqrvtTLyYIFo2HzdOPrRvdDj7HG4u2OYrYYS2d93AlO7YNZ9xl3Cru4O+6cJZD+9IlIyRQSZfTa+vxuOLnLWBT33hkQeauzKysSLauX48fH27AvOZ21iamsT0xl3YFTJJ+7yPqDqaw/mMqEpeBiNlGvkj8tLgWvJlWC8Pe8icnQUvxkphprER5cCZiMhdpvecyY2+hongHGMj2TOxkjzD+MMtpAFMW5SwgFLREpufzDYehC+PJeY9LvZ73hro+gbm9nV1YkTCYTNUN9qRnqy/23VMFms/H7qUzWJZ5i3YFU1iWmcvTMeRIOnyHh8Bk+XHEAkwmiK/oZI15Vy9G8ahBB3g4c9RDHOrkHZvSD04nGmoR9PoFaXYu2hvI1oe8n8EVf2PI5hNaHWx4p2hqKMQUtESnZPAPh/rkw+yHjdvOvhxhzRVr8w9mVFTmTyURkeW8iy3vTv1llAI6czrw02mWMciWmZPDbsTR+O5bG1NUHAagV6pMXulpUDSLEz8OJ70IKbF+88f/7xbMQUNlYFDo02jm11IiFzv+Dn/8FPz0LwbWgekfn1FLMKGiJSMnn6mncjbjgKdg4BX58Gs4dh05jyvwljPBAL8IDvbircTgASWkXjOB1adRrb3I6e5KMx2drfwegannvvEuNzauWo1KAp2OLzDyF7/kjpbYRrUOsnww//h/YciHiFuj/OfgEO7emlnGQvMPo2/X1YBi2FMpVd25NxYCCloiUDmYL9Hgb/CrCkpdg1Ttw7gTc/t7NNWgsZUL9POgVE0avmDAATqVfZMNB4zLjugOp7DyRRmJKBokpGczcYNzNGR7oSfOqQdxyadSrSjkvTPYIsMm7YO1EXLbOomPuRWyTPzXmFjXoZ4RnuVpuDiz8P9jwsfF1zL1GT6viMAHdZIKe70DKXjiyHr68x1jqx8Pf2ZU5lYKWiJQeJhO0fQp8KhjL9Wz9EjJOQt/p4O7j7OqKpXI+7txWryK31asIwNnz2Ww8aEyuX5uYyvajZzly+jxHTh9lzuajAIT6udO8ajlaVA3ilmpBVA/2KXjwstngwFKjDcG+xQCYACsWzCd3wfcjIP4/xlIxzR4C31BHvO2S6fxpY6TowDLABLEvQuvHi9eorYu7Mbo2uQOk7DEu6d870/iHUBmloCUipU/j+8EnBL4aZPwyn94TBnzt/EsrJYC/pyudokLpFGUEnPSLOWz+/TTrEk+xPjGVrYfPkpR2ke+3HuP7rccAKOftRvOqQXkT7OtU8MVsvuKXf85F2PY1rHnfWCcPABPU6UFO80f4acthbgtJxrLxYzh7GFa8AavHQb0+cMujULFB0f0QiqNT+2FGfzi1F1y94K7JENXT2VVdm2+osUzPJ7fB3p+N4Nz5v86uymkUtESkdKrVFQbPN+6EOrYFPuli9NoKqubsykoUH3cX2tYKpm0tI6ReyM5ly6EzecFr86HTnMrI4sftJ/hx+wkA/Dxc8oJXq4omoo58bQSojEsNLV29odF9xp1pQdWwZWeTsy0V6y1xWFoNh13fG4HsyHrYOsN4RLYxLivW6lr2RkcSV8Cs++HCGfCrZIwQFffgGdYI7pgIsx+E1eMhpC7E9Hd2VU6hoCUipVd400u9tu6C1APGsiQDv4bges6urMTycLXQsno5WlYvB0BWjpVtR8+w9lI7iU0HU0m7kEPiri102LuQGpYVWEzZAKS7hXCm/hBC2j+Cm2/QtU9gcTHac9TtDUc2wtr34bd5Ro+ogyuNJWVueRQaDiwbl4M3TjUakVpzoFJTuGdGybmcWr+PMTl+5dvw3T+hXA0Ib+LsqoqcgpaIlG7la8KDi+CLPnBiG0zrienuac6uqtRwczHTpIrRCDWuvY3c/cvJWD4ev8NL8vbZZo1kck4PFlxoQc5qFzzWr6NRROCluxqDqF/xOoEpvKnRF6rzf2H9R8Zi4qcTjbtKl7wMTR6A5v+AgIiiebNFKTcHfv43rJtkfF2vD9wxoeTdJNDh38ZND7t/gJkD4OGl4Bfm7KqKlIKWiJR+vhVg8AJjiZDE5Vhm3UNkpfsguz24lu07ouwiJwu2z4Y1E7EkbcMPABPU7o71lsewuNWn8cFUsi91sD+VkcWaA6dYc8BYNsjVYqKyl4UDnvvpEFWB+pX8sfx5jpd/uBG22j5t3OCwdhKk7odf3jMuMUb1MloLRDR3xru3vwtn4ZsHYd8i4+sO/4a2o4vXpPeCMpvhrg/h485wcifMHAhDFpS8wHgTTDabGpc4S1paGv7+/pw9exY/Pz+7Hjs7O5sFCxbQvXt3XF11a7uz6fMoJnKyYN6jsP0bAGwuHpiqdYA63aHWbcYEeim4zFSjb9n6jyHdmJ+Fq5dxWe+WR6/ZQ8lms7H/ZDprD6Tm9fNKSruYb58AL1da1yhP25rlaVsrmIr+V/xStlqNSdZrJxrzly6r1NQ4b/QdJbelR2qi0Rbh5C5w8YTeH0DdO4u0BIf8fZWaaNyJeP401O9nrOBQEoPjJYX5/a0RLREpO1zc4K7J5JavzcVfPsQrKwX2/Gg8MBkjIrW7Qe0eRmdrubaUvcbcqYQvIee8sc23IjR/GJoMBq/rzL/C6F5fI8SXGiG+3Hdp2aD9SWlM/m45pz0qsuZAKmcys/nh1+P88OtxAGqE+NC2ZjBta5WnRdVyeLpZoPZtxuPEdmOEa9tXcHSjMfl60QuXahlkrBxQUvz+izHqmnnK+HneMwMqNXZ2VfYRVNVoKvzpncZnFVoXbh3p7KqKhIKWiJQtZjPW1k+w6Ewtujetguv+RbDrBzieYCyKe3gdLH7RmLhbuzvU6QHhzcrenW5XstmMyehrJsKehX9sr9AAWg43Jq+7FH7NRJPJRJVyXrQKtdG9e0NMZgtbj5xhxZ4UVuw9ydbDZ9iXnM6+5HQ+WZ2Im4uZ5pFBtLk02lWnQl1Md06E2DGwYYrRyDPtKCweA8tfh4YDoMWjUL6GHX8YDrDlC6P3mzUbKjaEe78sfXOZqraFbq8bk/sXv2gsDF/U6zI6gYKWiJRNJhOE1oPwRtDuaTh7FHYvMB6JK+HUPvjlXePhVd64tFinO1TrAG5ezq6+6ORkwW9zYM0E42aCy2p1M+ZFRd5q10tALpY/Jtc/0bkWZzOzWb0/hRV7TrJiz0mOnb3Aqn0prNqXwqs/7iLY190IXTWDubXZKMrf+oRxafhyv64NHxsBrFZXoz1E1bbF65KVNdcIHb+8a3wdfQfc+UHp/X+s2UOQ9BtsmmrMQxsWD8G1nV2VQyloiYgA+FeC5sOMx4U0o9Hp7gWw52fITIGEz42HiydU72BcYqzVrfQ2Qc1MNX4Zrvvoj/lXLp7GCNEtjxXZCJG/lyvd61eke/2Kl+Z3ZbByrxG61h5I5eS5i8zZ/EfX+nqV/GhTsyltu3xHU9s2XDd8YIzAXX6E1jPmcdXrA65OXjz74jmY87Dx/xkYk/3bP2tMIC+tTCbo9oZx+fn3VUYT1mFL/vJyc0mnoCUiciUPP6h3l/HIzYbfV8PuH2HXAjh76I+Rr7x5XZcuMZav6ezKb96p/ZfmX82A7Exjm08FI4A2HerUX4jG/C4faoT4MKR1VS7m5LLp4GlW7DVGvHYcT2P7UeMxadl+vNwstKw2mh6tHqVT2hz8dn+NKWk7fBtnjCI1e8h4T864CeLMIfjyXkjaDhZ3o7lng75FX4czuLhBv+nG5PjTicayQvfNMXqolUKl812JiNiLxRWqtTcet71m/GLctcDoC3R865/mdY0pufO6bDYjTK6ZaARKLt2MHlofWg2Hunfd0PwrR3N3sdCqRnla1SjPM93qkHzuAqv3pbBiTwor954kJT2L+F3JxO8C6EpUQCdGlFtD+7Nz8cw4ActeNZpp1u8HLR8zJmgXhcPrjZ5SGSfBO8SYjxXetGjOXVx4l4d7vjSaCCcuh5//ZczfKoUUtERECspkggr1jUf7//v7eV21bzPuYKzWvnjOucnNht/mGvOvjm/9Y3ut2y7Nv2pTvOYz/Y0QXw96Nwqnd6NwrFYbO0+ksfLSaNfGg6fZecaFR8+0wYWW9HDZQJzHz9TK2f3HZeGq7YzLojW7OO7y3dZZ8N1wyM0yguy9X5bOhqsFUaGe0WNr1n2w7gMIiTbuFC1lFLRERG5UvnldZy/N6/rxj3ldWz43Hnnzui7363LyvK7zp40u6+s+gnPGwtC4ePxxh14paG1hNpuoG+ZP3TB/HmlXncysHNYdSGXFpfld355sybfpt9DYtJehLj/SzbIeS+JySFxOdmB1XFs+avw83LztU5DVCktfMkbQwAjgd31UNpYR+itRvaDDv2Dpy/DDk1C+FlRp6eyq7EpBS0TEHjz8od7dxuPyvK5dl0a7zh6+Yl5XC+MOxtrdi3Ze16n9xsjBls//NP8q1AiKTYaCd7miq6WIebm50KFOCB3qGPOxjp45z8o9J1mxtyLP7Y3m1QvHGeTyM/dYluB3ej8sGM35n/5Dcs17CIn9J57lq9z4ybMyYO4/YOf3xte3PgEdXyjdk94Lo+1Txp2IO+YZo1sPLytVo3wKWiIi9vbneV3dXr/GvK61xmPRC1Cu5qXQ1cOYp2PveV02GxxaY8y/2vUDf8y/qmdcHqx3N7i42/ecJUClAE/uaV6Ze5pXJtdqu9S7qxWP7h5KzWPfMdiykEiSqLJrMjk7p/CLV1tORA+ldpP2RFXww2wu4CXVs0dh5r3G525xg17vQsN7HfvmShqTCe5831j4/cSvxk0CD/5kv9FEJ1PQEhFxpKvmdR0xLi/mzevaC6vHGw/vYKPfkz3mdeVmw45vjflXx7b8sb1mFyNgVW1XouZfOZLFbKJx5UAaVw6E2FqcPd+ONfueZsXG76h36HMa8xutzi+DTcvYuKEWz7r0IqdWd26tXYFbawQT7HudoHp0E3w5wGiP4VUe7vkCKt9SpO+txHDzNjrhT+4ASdtg7iPQd3qpGPVT0BIRKUr+4VfP69q1APYuMu5Cyzevq+Mf6zB6ly/Y8c+fgc3TYd2HRod0MOZfxdxjTPQu5c0h7cHf05Xb6odD/cew2R7lyM51XFg5gcjjP9LUvIem1rc5snM607Z1ZUxue8IrVqRNrfK0qxlMk8hA3F0ssH2Osa5mzgUIjoIBsyDwJi4/lgUBEdD/c5jWE3Z+ByveNP5xUsIpaImIOMvfzuv6wXjkm9fV49rNQlMTjflXmz+D7Axjm3ewseZf06EFD2qSj8lkIjz6Foi+Bc6dIHfdx1g3TCH8Ygr/dv2CkS6z+fpkO6aeuI0Pl4fiZjHxnPd3DM76EoDfg25lR4t3qJjuT4TrRYK83TBpJPH6Kt8CPd8x7sxc9oqxTE/07c6u6qYoaImIFAdXzus6se2PCfRXzusqX+uPxa9tVlg7EXbOJ2/+VUj0pflXxaD7eWniWwFL7L+xtHsSfv0K1k7C5+ROhrj8xCCXn1lhasr5HOiWtQGAyTndefXYAKzf7AX2AuDlZiE80JOIQC/jv0HGf8MDvYgI8sLf09WJb7CYaHy/MTl+3STjJoKgqsal9xJKQUtEpLgxmaBiA+PR/pk/5nXt+gEOroKUPcZj9fj8r6sRawSsah00/8qRXD2Nfk+NH4D9S2DtJMz7FtHetgEsYDW5sKn+vznn0507Tp/nyOlMDqeeJ+ncBTKzctmTlM6epPRrHtrXwyVfCIv4UwgLD/TE272M/Nru8hKc3AUHlhrz3B5eWmJHZcvIJyYiUoL91byunAsQ09+YfxUS5exKyxaTCWp0Mh4nd8PaSXB8K+bO/6VZ1TY0u2L3izm5HDtzgcOpmRw5fZ7DpzPznh85nUlKehbnLuSw43gaO46nXfOUQd5ueeErPOhSCLsUyioFeOLhWkJWI/g7FhfoOxUmdzTuRpx1PzzwbbFcoeDvKGiJiJQk+eZ15RiXDkvgL59SJ7g29Br3l7u4u1ioWt6bquWv3bbgfFauMfp1+lIQyxfIznP2fDapGVmkZmSx9cjZax4jxNf9T6Nh+S9PhgV44mopQXfxeQbCvTPh41g49AssGA29xpe40VoFLRGRkqqULsJbVnm6WagZ6kvNUN9rfj/tQjZHUs9fFcSOXBoZy8jKJfncRZLPXWTzoTNXvd5sgor+nlS6Yo5YRKAn4UFeVPDzwFLQ/mBFJbg23D0FZvQz7qatUN8Y2S1B9KdURESkBPDzcCU6zJXoML+rvmez2TiTmZ03+nWtkbGLOVaOnjnP0TPnWZ+YetUxXMwmwgI8CQ/wgHQzOVuP065OKOV9nNzQtlYX6Pwf40aQH//PuBmkWjvn1lQICloiIiIlnMlkItDbjUBvNxqEB1z1fZvNxsn0i3kh7MpLk0dPnyfHauNQaiaHUjMBM798sw3YRr1KfrStGUybmsE0qRKIm4sTLj+2GgFJO+DXmfD1IBi2BIKqFX0dN0BBS0REpJQzmUyE+HoQ4utBkyqBV30/12ojKc2YqP97SjoL1/7KMas/u06cY/vRNLYfTeP9ZfvxdrPQsno52tYygldkOa+i6QtmMhnzs07tvdRx/154cBF4XD26V9woaImIiJRxlkuXDcMCPGkc4Yf78QS6d2/J6Qu5rNqbwoo9J1m5N4VTGVks3pnM4p3JAEQEeeaNdrWqUQ4/Dwf2AXP1gP5fGMv0nNwFcx42ljWy9/qgdqagZUe9e/dm2bJldOrUiW+++cbZ5YiIiNyUEF8P7moczl2Nw7Fabew4nsaKvSdZuSeFjb+ncjj1PF+sO8QX6w5dWjMygDY1g2lbK5j6lfztP7ner6IRrj7pBnt+hCUvQewY+57DzhS07Ojxxx9n6NChTJ8+3dmliIiI2JXZbKJeJX/qVfLnsfY1yLiYw9oDp1h5acTrQEoGGw6eZsPB04xdtIcAL1durVHeGPGqVZ6K/p72KaRSE7hjAswZBqvGQmhdqN/HPsd2AAUtO2rfvj3Lli1zdhkiIiIO5+3uQqeoUDpFhQJwODUzb7Rr9f4UzmRmM//X48z/9TgAtUJ98ka7WlQNurnmqg36Gcv0rB4H38YZE+MrNbbDu7I/p3cue/XVV2nWrBm+vr6EhIRw5513snv3brueY8WKFfTq1YuwsDBMJhPz5s275n4TJ04kMjISDw8PWrRowfr16+1ah4iISGkVEeTFwBZV+OD+Jmx5vjPfPNKSEZ1q0jAiALMJ9iSlM2VVIoM+WU+D//zM/VPWMXnFAXafOIfNZiv8CTu9ADW7GqsjzBwI507Y/03ZgdOD1vLly4mLi2Pt2rUsWrSI7OxsunTpQkZGxjX3X716NdnZ2Vdt37FjB0lJSdd8TUZGBjExMUycOPG6dcyaNYtRo0YxZswYNm/eTExMDF27diU5OTlvn4YNG1KvXr2rHseOHSvkuxYRESm9XCxmmkYGMapzLebFtWbz852ZOKAx/ZtGUNHfg6wcKyv3pvDygp10HbeCW16NZ/TXW/lu6zFSM7IKdhKzBe7+GMrXhnPHjLCVfcGxb+wGOP3S4cKFC/N9PW3aNEJCQti0aRNt27bN9z2r1UpcXBw1a9Zk5syZWCzGsOPu3bvp2LEjo0aN4umnn77qHN26daNbt25/WcfYsWMZNmwYQ4YMAeCDDz7ghx9+4JNPPuGZZ54BICEh4UbfpoiISJkV4OVGjwYV6dGgIjabjX3J6ay4NLdrXeIpktIu8s2mI3yz6QgmE9Sv5E/bS5cZG1UOuP7SQR5+cO+XxpqIRzfC/JFw56RitUyP04PWlc6eNdZvCgoKuup7ZrOZBQsW0LZtWx544AE+++wzEhMT6dixI3feeec1Q1ZBZGVlsWnTJp599tl854qNjWXNmjU39kb+wsSJE5k4cSK5ubl2P7aIiEhxZjKZ8pYaevDWqlzIzmXjwdOs2HuSFXtOsuvEOX49cpZfj5xlwtJ9+Li75PXualuzPFXKXbFWZLnq0G86fHYXbP3SmBzf6p/OeXPXUKyCltVqZeTIkbRu3Zp69epdc5+wsDCWLFlCmzZtGDBgAGvWrCE2NpZJkybd8HlTUlLIzc0lNDQ03/bQ0FB27dpV4OPExsaydetWMjIyCA8P5+uvv6Zly5ZX7RcXF0dcXBxpaWn4+/vfcN0iIiIlnYerhVtrlufWmuV5rnsUSWkX8u5kXLUvhdSMLBbtSGLRDmN6UJVyXnmjXS2rl8PH3QWqtYfbXoUfnzaW6gmuAzU7O/eNXVKsglZcXBzbt29n1apVf7lf5cqV+eyzz2jXrh3VqlVjypQpRdOZ9m8sXrzY2SWIiIiUaKF+HvRpEk6fJkbvrt+OGb27lu85yebfT/P7qUw+O/U7n639HRezicZVAmlXK5g2NfpRv9F2TFs+hW+GwkPxEFzL2W/H+ZPhLxs+fDjz589n6dKlhIeH/+W+SUlJPPzww/Tq1YvMzEyeeOKJmzp3+fLlsVgsV02mT0pKokKFCjd1bBEREbkxZrOJ+uH+xHWowVf/aEnCmC5MfqAp999ShchyXuRYbaxPTOXNn3Zz+8RfaLG1G/s968PFNHJm3APnTzv7LTg/aNlsNoYPH87cuXNZsmQJVatW/cv9U1JS6NSpE1FRUcyZM4f4+HhmzZrF6NGjb7gGNzc3mjRpQnx8fN42q9VKfHz8NS/9iYiISNHzcXehc3Qo/7uzHsue6sCKpzrwvzvr0SU6FB93F5IzbfQ7/RhHbOVxOb2fjW/fxcfL9ji1ZqdfOoyLi2PGjBl8++23+Pr6cuKE0QfD398fT8/8XWStVivdunWjSpUqzJo1CxcXF6Kjo1m0aBEdO3akUqVK1xzdSk9PZ9++fXlfJyYmkpCQQFBQEJUrVwZg1KhRDBo0iKZNm9K8eXPGjRtHRkZG3l2IIiIiUrxULufF/eWqcP8tVcjOtbLl0BlW7DnJWztf4JXTT9E0ZzNZv70F7T9yWo1OD1qXJ7G3b98+3/apU6cyePDgfNvMZjOvvPIKbdq0wc3NLW97TEwMixcvJjg4+Jrn2LhxIx06dMj7etSoUQAMGjSIadOmAdC/f39OnjzJCy+8wIkTJ2jYsCELFy68aoK8iIiIFD+uFjPNqwbRvGoQdK3Nuc0+2L57kOqVK4HN5rSWD04PWoXtBtu587XvImjUqNF1X9O+ffsCnWf48OEMHz68UPWIiIhI8ePb+G6IqE+okyfEO32OloiIiIhD6K5DERERkdJLQUtERETEQRS0RERERBxEQUtERETEQRS0RERERBxEQUtERETEQRS0RERERBxEQUtERETEQRS0RERERBxEQUtERETEQRS0RERERBxEQUtERETEQRS0RERERBzExdkFlGU2mw2AtLQ0ux87OzubzMxM0tLScHV1tfvxpXD0eRQv+jyKF30exYs+j793+ff25d/jf0VBy4nOnTsHQEREhJMrERERkcI6d+4c/v7+f7mPyVaQOCYOYbVaOXbsGL6+vphMJrseOy0tjYiICA4fPoyfn59djy2Fp8+jeNHnUbzo8yhe9Hn8PZvNxrlz5wgLC8Ns/utZWBrRciKz2Ux4eLhDz+Hn56c/KMWIPo/iRZ9H8aLPo3jR5/HX/m4k6zJNhhcRERFxEAUtEREREQdR0Cql3N3dGTNmDO7u7s4uRdDnUdzo8yhe9HkUL/o87EuT4UVEREQcRCNaIiIiIg6ioCUiIiLiIApaIiIiIg6ioCUiIiLiIApapdDEiROJjIzEw8ODFi1asH79emeXVCa9+uqrNGvWDF9fX0JCQrjzzjvZvXu3s8uSS1577TVMJhMjR450dill2tGjR7nvvvsoV64cnp6e1K9fn40bNzq7rDIpNzeX559/nqpVq+Lp6Un16tX53//+V6D1/OT6FLRKmVmzZjFq1CjGjBnD5s2biYmJoWvXriQnJzu7tDJn+fLlxMXFsXbtWhYtWkR2djZdunQhIyPD2aWVeRs2bODDDz+kQYMGzi6lTDt9+jStW7fG1dWVH3/8kR07dvD2228TGBjo7NLKpNdff51JkyYxYcIEdu7cyeuvv84bb7zBe++95+zSSjS1dyhlWrRoQbNmzZgwYQJgrKcYERHBP//5T5555hknV1e2nTx5kpCQEJYvX07btm2dXU6ZlZ6eTuPGjXn//fd56aWXaNiwIePGjXN2WWXSM888w+rVq1m5cqWzSxGgZ8+ehIaGMmXKlLxtd999N56ennz++edOrKxk04hWKZKVlcWmTZuIjY3N22Y2m4mNjWXNmjVOrEwAzp49C0BQUJCTKynb4uLi6NGjR74/J+Ic3333HU2bNqVv376EhITQqFEjJk+e7OyyyqxWrVoRHx/Pnj17ANi6dSurVq2iW7duTq6sZNOi0qVISkoKubm5hIaG5tseGhrKrl27nFSVgDGyOHLkSFq3bk29evWcXU6ZNXPmTDZv3syGDRucXYoABw4cYNKkSYwaNYrnnnuODRs2MGLECNzc3Bg0aJCzyytznnnmGdLS0qhTpw4Wi4Xc3FxefvllBg4c6OzSSjQFLZEiEBcXx/bt21m1apWzSymzDh8+zOOPP86iRYvw8PBwdjmC8Q+Qpk2b8sorrwDQqFEjtm/fzgcffKCg5QRfffUVX3zxBTNmzKBu3bokJCQwcuRIwsLC9HncBAWtUqR8+fJYLBaSkpLybU9KSqJChQpOqkqGDx/O/PnzWbFiBeHh4c4up8zatGkTycnJNG7cOG9bbm4uK1asYMKECVy8eBGLxeLECsueihUrEh0dnW9bVFQUs2fPdlJFZdtTTz3FM888wz333ANA/fr1+f3333n11VcVtG6C5miVIm5ubjRp0oT4+Pi8bVarlfj4eFq2bOnEysomm83G8OHDmTt3LkuWLKFq1arOLqlM69SpE9u2bSMhISHv0bRpUwYOHEhCQoJClhO0bt36qpYne/bsoUqVKk6qqGzLzMzEbM4fCywWC1ar1UkVlQ4a0SplRo0axaBBg2jatCnNmzdn3LhxZGRkMGTIEGeXVubExcUxY8YMvv32W3x9fTlx4gQA/v7+eHp6Orm6ssfX1/eq+XHe3t6UK1dO8+ac5IknnqBVq1a88sor9OvXj/Xr1/PRRx/x0UcfObu0MqlXr168/PLLVK5cmbp167JlyxbGjh3L0KFDnV1aiab2DqXQhAkTePPNNzlx4gQNGzbk3XffpUWLFs4uq8wxmUzX3D516lQGDx5ctMXINbVv317tHZxs/vz5PPvss+zdu5eqVasyatQohg0b5uyyyqRz587x/PPPM3fuXJKTkwkLC+Pee+/lhRdewM3NzdnllVgKWiIiIiIOojlaIiIiIg6ioCUiIiLiIApaIiIiIg6ioCUiIiLiIApaIiIiIg6ioCUiIiLiIApaIiIiIg6ioCUi4kSRkZFqmCpSiiloiUiZMXjwYO68807A6Ao/cuTIIjv3tGnTCAgIuGr7hg0bePjhh4usDhEpWlrrUETkJmRlZd3U8iTBwcF2rEZEihuNaIlImTN48GCWL1/O+PHjMZlMmEwmDh48CMD27dvp1q0bPj4+hIaGcv/995OSkpL32vbt2zN8+HBGjhxJ+fLl6dq1KwBjx46lfv36eHt7ExERwWOPPUZ6ejoAy5YtY8iQIZw9ezbvfC+++CJw9aXDQ4cOcccdd+Dj44Ofnx/9+vUjKSkp7/svvvgiDRs25LPPPiMyMhJ/f3/uuecezp0759gfmojcEAUtESlzxo8fT8uWLRk2bBjHjx/n+PHjREREcObMGTp27EijRo3YuHEjCxcuJCkpiX79+uV7/fTp03Fzc2P16tV88MEHAJjNZt59911+++03pk+fzpIlS3j66acBaNWqFePGjcPPzy/vfKNHj76qLqvVyh133EFqairLly9n0aJFHDhwgP79++fbb//+/cybN4/58+czf/58li9fzmuvveagn5aI3AxdOhSRMsff3x83Nze8vLyoUKFC3vYJEybQqFEjXnnllbxtn3zyCREREezZs4datWoBULNmTd544418x/zzfK/IyEheeuklHnnkEd5//33c3Nzw9/fHZDLlO9+V4uPj2bZtG4mJiURERADw6aefUrduXTZs2ECzZs0AI5BNmzYNX19fAO6//37i4+N5+eWXb+4HIyJ2pxEtEZFLtm7dytKlS/Hx8cl71KlTBzBGkS5r0qTJVa9dvHgxnTp1olKlSvj6+nL//fdz6tQpMjMzC3z+nTt3EhERkReyAKKjowkICGDnzp152yIjI/NCFkDFihVJTk4u1HsVkaKhES0RkUvS09Pp1asXr7/++lXfq1ixYt5zb2/vfN87ePAgPXv25NFHH+Xll18mKCiIVatW8eCDD5KVlYWXl5dd63R1dc33tclkwmq12vUcImIfCloiUia5ubmRm5ubb1vjxo2ZPXs2kZGRuLgU/K/HTZs2YbVaefvttzGbjQsFX3311d+e70pRUVEcPnyYw4cP541q7dixgzNnzhAdHV3gekSk+NClQxEpkyIjI1m3bh0HDx4kJSUFq9VKXFwcqamp3HvvvWzYsIH9+/fz008/MWTIkL8MSTVq1CA7O5v33nuPAwcO8Nlnn+VNkv/z+dLT04mPjyclJeWalxRjY2OpX78+AwcOZPPmzaxfv54HHniAdu3a0bRpU7v/DETE8RS0RKRMGj16NBaLhejoaIKDgzl06BBhYWGsXr2a3NxcunTpQv369Rk5ciQBAQF5I1XXEhMTw9ixY3n99depV68eX3zxBa+++mq+fVq1asUjjzxC//79CQ4OvmoyPRiXAL/99lsCAwNp27YtsbGxVKtWjVmzZtn9/YtI0TDZbDabs4sQERERKY00oiUiIiLiIApaIiIiIg6ioCUiIiLiIApaIiIiIg6ioCUiIiLiIApaIiIiIg6ioCUiIiLiIApaIiIiIg6ioCUiIiLiIApaIiIiIg6ioCUiIiLiIApaIiIiIg7y/5Ss8KLrzqTlAAAAAElFTkSuQmCC",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAkgAAAHHCAYAAABEEKc/AAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjguMywgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy/H5lhTAAAACXBIWXMAAA9hAAAPYQGoP6dpAAB2YElEQVR4nO3dd3gUVd/G8e9ueoeQRgmEJggiVRBQbBRFY0MRUJoIjwoW8lpAQbCBFbGAPPoAFkBRwY5ICAIiIE1QpEgPLYFQEtI3u/P+sWQhDZKYZDfJ/bmuXGRnZ2d+uyeQmzNnzjEZhmEgIiIiIg5mZxcgIiIi4moUkERERETyUUASERERyUcBSURERCQfBSQRERGRfBSQRERERPJRQBIRERHJRwFJREREJB8FJBEREZF8FJBExKWYTCYmTpxY4tft378fk8nERx99VOY1iUj1o4AkIgV89NFHmEwmTCYTq1atKvC8YRhERkZiMpm45ZZbnFBh2Vi0aBEmk4k6depgs9mcXY6IuBAFJBEpkre3N/PmzSuwfcWKFRw6dAgvLy8nVFV25s6dS1RUFEePHmXZsmXOLkdEXIgCkogUqXfv3nz55Zfk5OTk2T5v3jzat29PRESEkyr799LS0vj222+JiYmhbdu2zJ0719klFSktLc3ZJYhUOwpIIlKk/v37c+LECWJjYx3bsrOz+eqrrxgwYEChr0lLS+P//u//iIyMxMvLi2bNmvHGG29gGEae/bKyshg9ejShoaEEBARw6623cujQoUKPefjwYe6//37Cw8Px8vKiZcuWzJo161+9t6+//pqMjAzuvvtu+vXrx8KFC8nMzCywX2ZmJhMnTuSSSy7B29ub2rVrc+edd7Jnzx7HPjabjbfffptWrVrh7e1NaGgoN954Ixs2bAAuPD4q/5iriRMnYjKZ2LZtGwMGDKBmzZpcddVVAPz5558MGTKERo0a4e3tTUREBPfffz8nTpwo9DMbNmwYderUwcvLi4YNG/LQQw+RnZ3N3r17MZlMvPXWWwVet3r1akwmE5999llJP1KRKsXd2QWIiOuKioqic+fOfPbZZ9x0000A/PTTTyQnJ9OvXz/eeeedPPsbhsGtt97KL7/8wrBhw2jTpg0///wzTz75JIcPH87zC/mBBx5gzpw5DBgwgC5durBs2TJuvvnmAjUkJiZy5ZVXYjKZGDVqFKGhofz0008MGzaMlJQUHn/88VK9t7lz53LdddcRERFBv379GDNmDN9//z133323Yx+r1cott9xCXFwc/fr147HHHuPMmTPExsaydetWGjduDMCwYcP46KOPuOmmm3jggQfIycnh119/Ze3atXTo0KFU9d199900bdqUSZMmOcJlbGwse/fuZejQoURERPD333/zwQcf8Pfff7N27VpMJhMAR44coWPHjpw+fZoRI0bQvHlzDh8+zFdffUV6ejqNGjWia9euzJ07l9GjRxf4XAICArjttttKVbdIlWGIiOQze/ZsAzDWr19vvPfee0ZAQICRnp5uGIZh3H333cZ1111nGIZhNGjQwLj55psdr/vmm28MwHjppZfyHO+uu+4yTCaTsXv3bsMwDGPz5s0GYDz88MN59hswYIABGBMmTHBsGzZsmFG7dm0jKSkpz779+vUzgoKCHHXt27fPAIzZs2df9P0lJiYa7u7uxocffujY1qVLF+O2227Ls9+sWbMMwJgyZUqBY9hsNsMwDGPZsmUGYDz66KNF7nOh2vK/3wkTJhiA0b9//wL75r7X83322WcGYKxcudKxbdCgQYbZbDbWr19fZE3//e9/DcDYvn2747ns7GwjJCTEGDx4cIHXiVQ3usQmIhfUt29fMjIy+OGHHzhz5gw//PBDkZfXFi1ahJubG48++mie7f/3f/+HYRj89NNPjv2AAvvl7w0yDIMFCxYQHR2NYRgkJSU5vnr16kVycjKbNm0q8Xv6/PPPMZvN9OnTx7Gtf//+/PTTT5w6dcqxbcGCBYSEhPDII48UOEZub82CBQswmUxMmDChyH1K48EHHyywzcfHx/F9ZmYmSUlJXHnllQCOz8Fms/HNN98QHR1daO9Vbk19+/bF29s7z9irn3/+maSkJO67775S1y1SVSggicgFhYaG0r17d+bNm8fChQuxWq3cddddhe574MAB6tSpQ0BAQJ7tl156qeP53D/NZrPjElWuZs2a5Xl8/PhxTp8+zQcffEBoaGier6FDhwJw7NixEr+nOXPm0LFjR06cOMHu3bvZvXs3bdu2JTs7my+//NKx3549e2jWrBnu7kWPRtizZw916tQhODi4xHVcSMOGDQtsO3nyJI899hjh4eH4+PgQGhrq2C85ORmwf2YpKSlcdtllFzx+jRo1iI6OznOX4ty5c6lbty7XX399Gb4TkcpJY5BE5KIGDBjA8OHDSUhI4KabbqJGjRoVct7cuYnuu+8+Bg8eXOg+l19+eYmOuWvXLtavXw9A06ZNCzw/d+5cRowYUcJKL6yoniSr1Vrka87vLcrVt29fVq9ezZNPPkmbNm3w9/fHZrNx4403lmoep0GDBvHll1+yevVqWrVqxXfffcfDDz+M2az/O4soIInIRd1xxx385z//Ye3atcyfP7/I/Ro0aMDSpUs5c+ZMnl6kHTt2OJ7P/dNmszl6aHLt3Lkzz/Fy73CzWq107969TN7L3Llz8fDw4NNPP8XNzS3Pc6tWreKdd94hPj6e+vXr07hxY37//XcsFgseHh6FHq9x48b8/PPPnDx5sshepJo1awJw+vTpPNtze9SK49SpU8TFxfH888/z3HPPObbv2rUrz36hoaEEBgaydevWix7zxhtvJDQ0lLlz59KpUyfS09MZOHBgsWsSqcr03wQRuSh/f3/ef/99Jk6cSHR0dJH79e7dG6vVynvvvZdn+1tvvYXJZHLcCZf7Z/674KZOnZrnsZubG3369GHBggWF/sI/fvx4id/L3Llzufrqq7nnnnu466678nw9+eSTAI5b3Pv06UNSUlKB9wM47izr06cPhmHw/PPPF7lPYGAgISEhrFy5Ms/z06dPL3bduWHOyDddQv7PzGw2c/vtt/P99987phkorCYAd3d3+vfvzxdffMFHH31Eq1atStwjJ1JVqQdJRIqlqEtc54uOjua6667j2WefZf/+/bRu3ZolS5bw7bff8vjjjzvGHLVp04b+/fszffp0kpOT6dKlC3FxcezevbvAMV955RV++eUXOnXqxPDhw2nRogUnT55k06ZNLF26lJMnTxb7Pfz+++/s3r2bUaNGFfp83bp1adeuHXPnzuXpp59m0KBBfPLJJ8TExLBu3Tquvvpq0tLSWLp0KQ8//DC33XYb1113HQMHDuSdd95h165djstdv/76K9ddd53jXA888ACvvPIKDzzwAB06dGDlypX8888/xa49MDCQbt268dprr2GxWKhbty5Llixh3759BfadNGkSS5Ys4ZprrmHEiBFceumlHD16lC+//JJVq1bluUQ6aNAg3nnnHX755RdeffXVYtcjUuU57wY6EXFV59/mfyH5b/M3DMM4c+aMMXr0aKNOnTqGh4eH0bRpU+P111933F6eKyMjw3j00UeNWrVqGX5+fkZ0dLRx8ODBAre9G4b9tvyRI0cakZGRhoeHhxEREWHccMMNxgcffODYpzi3+T/yyCMGYOzZs6fIfSZOnGgAxpYtWwzDsN9a/+yzzxoNGzZ0nPuuu+7Kc4ycnBzj9ddfN5o3b254enoaoaGhxk033WRs3LjRsU96eroxbNgwIygoyAgICDD69u1rHDt2rMjb/I8fP16gtkOHDhl33HGHUaNGDSMoKMi4++67jSNHjhT6mR04cMAYNGiQERoaanh5eRmNGjUyRo4caWRlZRU4bsuWLQ2z2WwcOnSoyM9FpLoxGUa+/loREalW2rZtS3BwMHFxcc4uRcRlaAySiEg1tmHDBjZv3sygQYOcXYqIS1EPkohINbR161Y2btzIm2++SVJSEnv37sXb29vZZYm4DPUgiYhUQ1999RVDhw7FYrHw2WefKRyJ5KMeJBEREZF81IMkIiIiko8CkoiIiEg+miiylGw2G0eOHCEgIOBfrdgtIiIiFccwDM6cOUOdOnUuuO6gAlIpHTlyhMjISGeXISIiIqVw8OBB6tWrV+TzCkillLsQ58GDBwkMDCyz41osFpYsWULPnj2LXBxTKpbaxLWoPVyL2sO1qD0uLiUlhcjIyDwLahdGAamUci+rBQYGlnlA8vX1JTAwUD/cLkJt4lrUHq5F7eFa1B7Fd7HhMRqkLSIiIpKPApKIiIhIPgpIIiIiIvloDFI5s1qtWCyWYu9vsVhwd3cnMzMTq9VajpVJceW2ic1mc3YpIiJSQRSQyolhGCQkJHD69OkSvy4iIoKDBw9qfiUXkdsm+/bto1GjRnh6ejq7JBERKWcKSOUkNxyFhYXh6+tb7LBjs9lITU3F39//ghNYScWx2WycOXOGlJQUjh49Sv369RVeRUSqOAWkcmC1Wh3hqFatWiV6rc1mIzs7G29vbwUkF5HbJqGhoRw9epScnBzdPisiUsXpN3A5yB1z5Ovr6+RKpCzlhiKNDRMRqfoUkMqRLsNULWpPEZHqQwFJREREJB8FJCk3UVFRTJ061dlliIiIlJgGaUse1157LW3atCmTYLN+/Xr8/Pz+fVEiIiIVTD1IUiKGYZCTk1OsfUNDQzVQXURESizTYmXbkRQyLc67KUYBSRyGDBnCihUrePvttzGZTJhMJj766CNMJhM//fQT7du3x8vLi1WrVrFnzx5uu+02wsPD8ff354orrmDp0qV5jpf/EpvJZOJ///sfd9xxB76+vjRt2pTvvvuugt+liIi4itwg9O3mw7zx805GfLKB699YTovnFtP7nV/ZdjTFabXpElsFMAyDjGKmYJvNRka2FffsnDKZB8nHw63Yd1+9/fbb/PPPP1x22WW88MILAPz9998AjBkzhjfeeINGjRpRs2ZNDh48SO/evXn55Zfx8vLik08+ITo6mp07d1K/fv0iz/H888/z2muv8frrr/Puu+9y7733cuDAAYKDg//1exUREdeUabGy93gau46dYVdiKv8knmH3sVT2n0jDZhT+mkBvd06mZldsoedRQKoAGRYrLZ772Snn3vZCL3w9i9fMQUFBeHp64uvrS0REBAA7duwA4IUXXqBHjx6OfYODg2ndurXj8YsvvsjXX3/Nd999x6hRo4o8x5AhQ+jfvz8AkyZN4p133mHdunXceOONJX5vIiLiWkobhC4JD6BpeABNw/y5JDyAS8L9CQ3wcur0KgpIUiwdOnTI8zg1NZWJEyfy448/OmaXzsjIID4+/oLHufzyyx3f+/n5ERgYyLFjx8qlZhERKR+ZFiv7ktL4J9EehHIDUWUMQkVRQKoAPh5ubHuhV7H2tdlsnEk5Q0BgQJldYisL+e9Ge+KJJ4iNjeWNN96gSZMm+Pj4cNddd5GdfeHu0PxLdJhMJmw2W5nUKCIiZctpQWjPMtj4MfT5H7g5Z2knBaQKYDKZin2Zy2azkePphq+nu1PWYvP09CzWUhq//fYbQ4YM4Y477gDsPUr79+8v5+pERKQ8uFSP0N/fwIIHwGaBuu2g62OlP9a/oIAkeURFRfH777+zf/9+/P39i+zdadq0KQsXLiQ6OhqTycT48ePVEyQi4uJcKggVZuNH8MNoMGzQ8g7o9FDZHr8EFJAkjyeeeILBgwfTokULMjIymD17dqH7TZkyhfvvv58uXboQEhLC008/TUqK827HFBGRc7IsVnYnZbhuECrMqqmwdIL9+/ZD4OYpYC6bYSKloYAkeVxyySWsWbMmz7YhQ4YU2C8qKoply5bl2TZy5Mg8j/NfcjOMgn8rT58+Xao6RUSqE5vN4ExmDqfSszmVns3pdIvjz9Pp2Zw6+/hUWja7Drsxem1cMYKQP03DAhzfhzlrsLRhwNKJ8NtU++OrYuCG58DJA7cVkERERCpI7rx45wecU2cDTvJ5QSc3+OQ+n5xhKTLwFGQPFi4XhApjs9ovqW362P64x4vQ9VHn1nSWApKIiEgpWKw2TqdbSM44G2zSzuvZyTjbs5N2LuDkBqHsnNKP1/TzdKOGryc1fD2ome/PGr6eBHqZ2bdtCwNuuY66wf6uE4QKk5MFC0fAtm/AZIbot6HdIGdX5aCAJCIi1V5GtpVjZzI5la/nptCenYxsTqdZOJNVvHUpC+PhZrIHHZ/Cg07NQv4M8vXAy/3CY3IsFguLjmwmPNDbtcNRVirMvw/2/gJunvbb+Vvc5uyq8lBAEhGRauF0ejYHTqRz4GQ6B5LSOHAynfgT6ew/kcaxM1mlPm6Qj0eeYOMIOj6e1PQ7L+j4nA1Cfp74eRZ/GagqJ/0kzOsLh9aDhx/0mwuNr3N2VQUoIImISJVgGAbHzmSxP1/4iT+ZzoET6SRnWC74em8PM8G+ngTlCzqF9ezkbg/y8cDNXE2DTmmkHIU5d8KxbeBdA+5bAPU6XPRlzqCAJCJSUQwDMk5BymH7LwqfmhB2KXj5O7uySiPHauPw6Qx7T9CJtHM9QmeDUKblwuN7wgK8aFDLl/rBfkTV8qV+LV8a1PIjKsBGjcAgp95WXuWd3Auf3A6nD0BAbRj4tf3n30UpIImIlAXDgMxke/hJPnw2BOX7PuUIWNILvrZmQwhvCeGX2f+MuAxqRIETZtN3BRnZ1rO9PvbQs/9sEIo/mc6hUxlYL3A7l9kEdWv6EFXLj/rBvjQ4G4DsocgXXzcDTuyGxL/tvRjbtsGxv+F0PPjWguh34NJbKvDdVhOJf8Ond0Bqov3nfdA3UDPK2VVdkAKSiEhxZKZcOPwkHwZLWvGO5VsLAupA2nFITYBT++xfO344t4+HH4S3yBucwlqAT41yeXsVLTndwoGTaew/kU58bk/QiXQOnEwjMeXC44G83M2Fhp+oWn7UremDh5vZHlhTDkPi2QC0Y5s9ECX9A9Yi1oxMPwHz74UO90PPl8HTtxzeeTUU/zvMu9v+H4jwy+C+hRAQ7uyqLkoBSUQkK7Xo0JPb85NVzJnifWpCYD0IqguBdSCwrv0r6OyfgXXAw+fc/mlJ9v9dO762wrHt9rB1aL3963xBkWdDU8tz4Sm4Mbi51j/nueOB8l8Kiz9hD0UXGw8U4O1u7wWq5UtULV8aBOd+70dYgBfm88f9ZCbbP7N9f8Pv2871DmUmF35wT3/7pZ2wFueCZ0hTWPMerH4XNsyCA6uhz0x7b56U3u6lMH+gvec08koYML/ShHzX+hsllV5UVBSPP/44jz/+OGBfqPfrr7/m9ttvL3T//fv307BhQ/744w/atGlT6vOW1XGkCspOv0j4OVz0L9L8vIPyhZ983wfWKXmvg18INLrG/pXLmgMn99jD0vnhKfngua9/Fp/b380LwppDeKu8wcmvVslqKSGbzeDQqQx2nDZxat1BDp/OPNsjZL8clmG58MLXoQFe9nFAwX5ne4PO9ggF+1LD16PgXV452XBiFxw42yuUeLZXKPlg4ScwudmDT1gLe29cWEv7n0H1C7982fMlaHQdfP0gHN8BH14PPV6ATv9x+qzOldLWhfZ5jmwWaNId+n5aqXrlFJCkXB09epSaNWuW6TGHDBnC6dOn+eabbxzbIiMjOXr0KCEhIWV6LnFxlgx7744j8ByyPz4//GScKt6xvALP6+kpIvxU1GBqN3cIbWb/uqzPue0Zp+2BILenKfFsSLCkwdEt9q/z+UfkDUzhLSHkEnD3LFE5OVYb8SfT2XUsld1nv3YdO8OeY2lnQ5AbbN9e4HW544EanBeA6gf7ERVydjyQZxG/ggzDHnoS8wWhpH/AVsTcQ4F1CwahkEvA3atE75UmN8BDq+HbkbDrZ1j8NOyJg9umg39oyY5VnW2YbZ8hGwNa3gl3/LfEP3fOpoAk5SoiIqJCzuPm5lZh5xInOr4Tt+Wvcs3eDbjvHG0fM1Icnv7nLm8F1T0XeM7/3juwXEsvEz41oEEX+1cumw1O7897iS5hq31MU2qC/WtP3Ln9ze4Q0uxccIq4zB6e/MPJstrYl5RmD0CJ58LQvqQ0sq2F3x3m4WYi2NNGy/phRIX4O8JPg1p+1K3hg6f7RQaaZ5w6F4ByL40d2170JU2vwLyXx8Jb2h/7lOF/xPxD7ZeC1n0AS8bDriXwfhe4Y4Y9QEnRDANWvQVxz9sfd7gfer9RKe8OVEAShw8++ICJEydy6NAhzOd1P992223UqlWLZ599lpiYGNauXUtaWhqXXnopkydPpnv37kUeM/8ltnXr1vGf//yH7du3c9lll/Hss8/m2d9qtTJixAiWLVtGQkIC9evX5+GHH+axxx4DYOLEiXz88ceOYwP88ssvREVFFbjEtmLFCp588km2bNlCcHAwgwcP5qWXXsLd3f5jf+2113L55Zfj7e3N//73Pzw9PXnwwQeZOHFiWXycUpZysuDXKbBqCmZrNjXOf87DN98Yn0K+9wqsupdIzGYIbmT/ujT63PasVHvQyH+ZLivZ3itz7G/469zupwlkmy2S7bb67DDsf+4y6pGF/X/9Ph5uNAnzd3w1Pftn7QAPlvy8mN692+Lh4VF0nTlZ9h6g/L1CKYeLeF/u9h6gApfHIiumLU0m+6W1Bl1hwTD7Jbc5d0LnUfaFVEvaM1UdGAbEPger37E/vvoJuH5cpf275/SANG3aNF5//XUSEhJo3bo17777Lh07dix0X4vFwuTJk/n44485fPgwzZo149VXX+XGG2907DN58mQWLlzIjh078PHxoUuXLrz66qs0a9bMsc+1117LihUr8hz7P//5DzNmzCifN2kYhd/aWxibzb5vtlvZ3OLr4VvsH867776bRx55hF9++YUbbrD/L+nkyZMsXryYRYsWkZqaSu/evXn55Zfx8vLik08+ITo6mp07d1K/fv2LHj81NZVbbrmFHj16MGfOHPbt2+cIPrlsNhv16tXjyy+/pFatWqxevZoRI0ZQu3Zt+vbtyxNPPMH27dtJSUlh9uzZAAQHB3PkyJE8xzl8+DC9e/dmyJAhfPLJJ+zYsYPhw4fj7e2dJwB9/PHHxMTE8Pvvv7NmzRqGDBlC165d6dGjR7E+M6kA+3+DHx63/3IFbI27s85oRfvrb8ejVgP7ZHOV9B/gcuXlT3JIG3bbmrDH6M4uzrCLM5xJ3E/QmX9obornUnM8zU0HaWQ6Qg1TCl3Mf9PF/LfjEDbMZARGYQq/DO96l2OOuAzC6+cJKRZLvsHWNhskxxdyeWwXGEWMSQqKLBiEajV1jUsyEZfBiOWwZBys/599IPe+lXDXLPv4JrGzWeH7x+CPT+2Pe74MXUY5t6Z/yakBaf78+cTExDBjxgw6derE1KlT6dWrFzt37iQsLKzA/uPGjWPOnDl8+OGHNG/enJ9//pk77riD1atX07ZtW8DeazBy5EiuuOIKcnJyeOaZZ+jZsyfbtm3Dz8/Pcazhw4fzwgsvOB77+pbjwDFLOkyqU6xdzZD3f8f/1jNHwNPv4vsBNWvW5KabbmLevHmOgPTVV18REhLCddddh9lspnXr1o79X3zxRb7++mu+++47Ro26+F+EefPmYbPZmDlzJt7e3rRs2ZJDhw7x0EMPOfbx8PDg+eefdzxu2LAha9as4YsvvqBv3774+/vj4+NDVlbWBS+pTZ8+ncjISN577z1MJhPNmzfnyJEjPP300zz33HOOHrLLL7+cCRMmANC0aVPee+894uLiFJBcQcYpiJ1wbpVvvzC46VWsl9xC4k8/2S+tXKjHoho5kZrFrmOp7DqWyp6z44N2JaYWsXyGP9COLX5X2nuCwv1pFuxOK69EGtn2E3B6B6Zjf0PCVswZJ/FL2Qspe2HXd+cO4RXkuLxlDmlGw+ObcfsxFpJ22HutslMLL9Qr6GwIynd5zDuoPD6WsuPhAze/CY2vt49NSvgT/tsNbnoV2g5UQM/JggUPwPbv7IvO3voutL3P2VX9a04NSFOmTGH48OEMHToUgBkzZvDjjz8ya9YsxowZU2D/Tz/9lGeffZbevXsD8NBDD7F06VLefPNN5syZA8DixYvzvOajjz4iLCyMjRs30q1bN8d2X19fjVkpxL333svw4cOZPn06Xl5ezJ07l379+mE2m0lNTWXixIn8+OOPHD16lJycHDIyMoiPjy/Wsbdv3+64pJWrc+fOBfabNm0as2bNIj4+noyMDLKzs0t8Z9r27dvp3LlznrtgunbtSmpqKocOHXL0eF1++eV5Xle7dm2OHTtWonNJGTMM+Hsh/DQG0s62RbvB0ON5+ziT/D0W1YRhGCSkZJ4bH3Q8ld2J9jB0Kr3ozyQi0Jum4edfGgugSZg/wX75e2fyzWhsGPZJ/fJfoju+036ZLn41xK/GDbgc4NB5rzV72AeY5+8VCqxbucNE85uhTlv4+j/2XqTvHrHfxh79dtmOgapMslLtc0ftXW5fdPauWXkv9VZiTgtI2dnZbNy4kbFjxzq2mc1munfvzpo1awp9TVZWVp5frgA+Pj6sWrWqyPMkJ9tv3w0ODs6zfe7cucyZM4eIiAiio6MZP358+fUiefjae3KKwWazkXLmDIEBAXnGAf2rc5dAdHQ0hmHw448/csUVV/Drr7/y1ltvAfDEE08QGxvLG2+8QZMmTfDx8eGuu+4iO7uISddK4fPPP+eJJ57gzTffpHPnzgQEBPD666/z+++/l9k5zpd/zITJZMJmu/BSBVKOTsfDj/9nHxQL9jEo0W/nHZRcxeXeOr/7uL0X6Pw7x1KLWD3eZILImr55xgY1CfOncZg/gd6l7GUzmSAgwv7V5Lxxhrm32p8dEG5L+JvE4ycIu+xa3GqfvVOuVhNwq6K9e4F1YOC3sPptWPYSbPsWDm2EPh9Wq59TwL7o7Ny74fAG+8Sm/edBo2udXVWZcVpASkpKwmq1Eh6edzbN8PBwduzYUehrevXqxZQpU+jWrRuNGzcmLi6OhQsXYrUWfl3bZrPx+OOP07VrVy677NxkXwMGDKBBgwbUqVOHP//8k6effpqdO3eycOHCIuvNysoiK+tcd3VKiv0OC4vFUuAavMViwTAMbDbbuV+27j4Uh2EY4GHF8PDFVhb/0zIM+1cxeXp6cscddzBnzhx27dpFs2bNaNOmDTabjd9++43Bgwdz2223AfYxRfv373e81/Pfw/mPcz+HZs2a8emnn5Kenu4IuqtXr86zz6pVq+jSpQsPPvig4/V79uxx7AP2UJOTk1PgHOcfp3nz5o6fjdxepFWrVhEQEECdOnUc+xdWe2Hbzn/OYrHg5lb57shwabYczOs/xLxiMiZLOoabJ7Yuj2Pr8ph9MOx5f8dy/74VGPtSyVisNuJPZrDneCq7j6Wx+3gqe46nsTcprcj1xNzMJhoE+9I41I8mYX40CfWncagfjUL88PEs/Gey7D8nEwRfYv+69A4sFgvrYmPpcVWPc//hsGGf+6Yq6zQKU2QX3L75D6ZT+zA+uhlb19HYrn7SPsDcSSrs70fKUdw/uwtT0k4Mn5pY75mPUbddpejhLe5n4/RB2iXx9ttvM3z4cJo3b47JZKJx48YMHTqUWbNmFbr/yJEj2bp1a4EephEjRji+b9WqFbVr1+aGG25gz549NG7cuNBjTZ48Oc/YmFxLliwp0PPk7u5OREQEqamppe5dOXPmTKleVxZuv/12+vXrx9atW+nbt68jDEZFRfHVV19x3XXXATBp0iRsNhvZ2dmOfWw2G5mZmY7HABkZGaSkpHDLLbcwbtw4hg4dyujRo4mPj+eNN94AIC0tjZSUFCIjI/nkk0/4+uuvadCgAfPnz2fdunU0aNDAccyIiAgWL17Mxo0bCQ4OJjAwkNTU1DzHue+++3j77bd58MEHGT58OLt372bChAk8/PDDjn1zcnLy1J67zWKx5NmWKy0tjYyMDFauXElOThFzsUiJBaXvp038LGpk7Acgya8ZW+oPITW1LiyJK/J1sbGxFVRh2TEM2HMG1h83s/mEiUxr4f8JcjcZhPlAuI9BhI9BhK/9+1BvcDcnA8mQDRyGA4fhQIW+i8JVxvYoC+6RY2hl+pT6J1fhtupNkv/4lo0NHiLdy7lzJpVne/hlJdJ596t4ZCeR4VGTNQ2e4syWBNiyqNzOWZbS04t305TTAlJISAhubm4kJibm2Z6YmFjk2KDQ0FC++eYbMjMzOXHiBHXq1GHMmDE0atSowL6jRo3ihx9+YOXKldSrV++CtXTq1AmA3bt3FxmQxo4dS0xMjONx7i/znj17EhiYd/6UzMxMDh48iL+/f4FLghdjGAZnzpwhICCg4CyyFeSWW24hODiYXbt2MWTIEMf7e/vtt3nggQfo1asXISEhPPXUU2RkZODp6enYx2w24+3tnecz8fHxITAwkMDAQL777jsefvhhrrnmGlq0aMGrr77K3XffjZ+fH4GBgTz66KNs376dYcOGYTKZ6NevHw8//DCLFy92HHPUqFGsXbuW66+/ntTUVOLi4oiKigJwHCcwMJAffviBp59+mquvvprg4GCGDRvGCy+84LjN393dPU/tuds8PDzybMttEz8/P3x8fOjWrVuJ21UKkZ2GeeWrmDf/F5NhxfAKxHrDRILa3Ec3U9GXly0WC7GxsfTo0ePCt5W7kPiT6Xyz+Qhfbz7KoVMZju2+nm723qBQPxqH+tMk1I8mYf7Uq+mDm7lyjNWpjO1R9vqQ8/cC3H56guC03XTfMxHrTW9gtOxz8ZeWsXJvj8StuH/2BKbsJIyaDXEfsICra1z8LmZXUth/gAtjMowSXH8pY506daJjx468++67gL33oX79+owaNarQQdr5WSwWLr30Uvr27cukSZMA+y+zRx55hK+//prly5fTtOnFb8P87bffuOqqq9iyZUuBQbtFSUlJISgoiOTk5EID0r59+2jYsGGJf5HabDZSUlIIDAwsmzFI8q/ltomnpycHDhwoVbtKPrti4YcY++3gYJ9p98ZXirWApcViYdGiRfTu3dulfyGfybSw6K+jLNh4mHX7Tzq2+3u507tVBH3a1eOKqOC8a4pVQpWlPSrEqQP2u7kOrbM/bj0Aer8GXgEVVkK5tkf8Wpjb1z5IP6KVfdFZ/4J3nLu6C/3+Pp9TL7HFxMQwePBgOnToQMeOHZk6dSppaWmOu9oGDRpE3bp1mTx5MgC///47hw8fpk2bNhw+fJiJEydis9l46qmnHMccOXIk8+bN49tvvyUgIICEhAQAgoKC8PHxYc+ePcybN4/evXtTq1Yt/vzzT0aPHk23bt2KHY5EpJRSj8HisbD1K/vjoEj77dOX9HJuXWXEajNYtTuJhZsO8fPfCY6xRCYTXNUkhD7t6tGrZUSR44WkkqvZAIb+BCtfg5Wvw5Z5EL8G7poJdds7u7p/Z9dSmH8f5GRA/c7Q//NKs+hsaTk1IN1zzz0cP36c5557joSEBNq0acPixYsdA7fj4+Pz9KJkZmYybtw49u7di7+/P7179+bTTz+lRo0ajn3ef/99wD4Z5Plmz57NkCFD8PT0ZOnSpY4wFhkZSZ8+fRg3bly5v1+Rassw7BPILRkPmaftc6V0egiue6bi1jcrR7sSz/DVpkN888dhElPO3czRONSPPu3rcUfbutQOKt6NGlLJubnbf64bXQsLhtuXfJnZ0z6jdJfHymYC4Ir211f2qQ1sOdCkB/T9pFItOltaTh+kPWrUqCInGVy+fHmex9dccw3btm274PEudsUwMjKywCzaIlKOknbB94/DgbM3S0RcDre+Y59PphI7lZbNd1uOsGDTIf48lOzYXsPXg1tb16FPu3pcXi/IaWMJxckadIGHVtl/9rd9A0snwp5l9kVbA4s3cbBLWD/TPvUGBlx2F9z+vmvMcF4BnB6QRKSKysmG36baLzVYs+1zcl33jL3nyK1y/tOTnWNj+c5jLNh0iGU7jmGx2v9D5m42cW2zMO5qX5frmofh5a5LaIJ98si7P4I/5sBPT9knl3y/K9z2nn3SSVdmGPDrm7DsRfvjKx6Am16vnD1gpVQ5/5WqJJw4/l3KgdqzBOLX2tdlOn52TrMm3e1jjWpGObWs0jAMg62HU1iw6RDfbTnCybRzU3e0rBNIn3b1uLVNHUL8tXipFMJkgnYDof6V9kVvj26BzwdAh2HQ62X7MiauxjDsa8+tec/+uNuTcN2zlXsW9FJQQCoHuXcOpKen4+Pjgj/8Uiq5k4tpksgLyDhtv5Sw0b6QML4h9vWqLutT6f5xPZaSydd/HGbBpkP8k3hubbHQAC9ub1OHPu3r0Tyi6DtgRPIIaQrDYu09MqvfhQ0z4cBv0GemfUFcV2HNsf/nZrN9+S56TYbODzu3JidRQCoHbm5u1KhRw7Gml6+vb7HHIeROvJiZmanb/F2EzWYjKyuLlJQUfH19HfMoyXkMw77kwk9P2dfvAvsinj1eAN/gC7/WhWRarCzZlsiCjYf4dddxbGc7DT3dzfRsEU6f9vW4ukkI7m76uyml4O4FPV+CRtfB1w/ae1g/vB56vggdRzj/PxGWTHsv144fwORmvxTYZoBza3Ii/UtfTnInuyzpwqeGYZCRkYGPj48Gd7qI3Dbx8/Ojdu3aapf8kg/Bj0/APz/ZH9dqArdMhYZXO7Ws4jIMg40HTrFg0yF++PMoZzLPzZLevkFN+rSrx82X1ybIp5rP8SNlp8kN8NBq+HYk7PrZ/h+L3XFw+3TwC3FOTVln7Jf+9q0ENy+4e7brj5MqZwpI5cRkMlG7dm3CwsJKtCaOxWJh5cqVdOvWTZOuuQiLxcKKFSvo0aMHnp7V4+6NYrFZYd0H9gU7s1PtK7hfNRqu/j/wcP2JNA+eTGfhpsMs/OMQB06cW3qgbg0f7mxXlzvb1aNhiJ8TK5QqzT8UBsyHdR/ax/vs+hne72K/S6zJDRVbS/pJmNMHjmwCT3/o/xk07FaxNbggBaRy5ubmVqIxK25ubuTk5ODt7a2A5CLc3NywWq265Hm+o3/axykc2WR/HHklRL8NYc2dW9dFpGblnJ3d+hC/7zs3u7Wvpxs3XVabPu3rcmXDWpV+dmupJEwm6DTCPiXAgmH2S25z7oQuj8D1z1XM7fTJh+HTOyBpJ/gEw31fVf5JLcuIApKIFF92Oqx4BVa/B4YVvAKhx/PQbojL3v5rtRms2XOCBZsOsXhrAhkWK2D/3dSlcS3H7NZ+XvrnUJwk4jIYsdzek7T+f/ZB3PtW2gdwh1x8uaxSO7EHPrndvuRPQB0Y9A2ENiu/81Uy+hdBRIpndxz8MBpOn107vsVtcOOrEFjbuXUVYc/xVBZsPMTXfxzmaHKmY3ujEPvs1re3rUvdGrrLVFyEh499KozG19vHJh3dAv/tBje9Bm3vK/sB3Ef/tPdWpR2H4Mb2cFTJFp0tbwpIInJhqcfh52fgry/sjwPrQu83oHlv59ZViNPp2Xy/5QhfbTrMloOnHdsDvd2Jbm2/Nb9tZA0NtBfX1fxm+yzzX//H3ov03SjYvRSip9onniwLB1bDvHsgK+XsorNf28dESR4KSCJSOMOAzfNgybOQcQowQacH4fpnK3R18ouxWG2s2HmcBZsOEbf9GNlW+wKxbmYT11wSSp929bjh0jC8PTR/lVQSgXVg4Lew+m37TRDbvoHDG+HOD+zjlf6Nf5bAFwMhJxPqd4EBn4N3UJmUXdUoIIlIQSf22Adh7//V/ji8lX0Qdj3XGby57WgK325J5Lsth0lKPTe7dfOIAO5qX4/b2tQlNECzW0slZTbb7wpt2A0WPAAn98JHN9tnte72VOmW6zl/0dlLbrQvg+KKM3m7CAUkETknJ9v+v9YVr4M1C9x94Nox0HkkuDnvrkqbzeDw6Qz2HE9l66HTzNvixpE1ax3Ph/h7clubuvRpV48WdTS7tVQhddvDf1bCT0/D5rmw4lXYuxzu/BBqNij+cdZ9CIueBAxo1dc+55IT/05XBgpIImJ3cB189ygc325/3Og6uOUtCG5YYSVkZFvZm5TKnuNp7DmWyp7j9u/3Hk8lK8d23p4mPNxM9GgRTp929eh2SSgemt1aqiqvAHugaXy9/UaJg7/DjKvsfz9b3XXh1xoGrHwDfnnJ/rjjCPvNFS5616krUUASqe4ykyHuBVg/EzDAtxbc+Aq0urtclj4wDIPjZ7LYfTxvENp7PI3DpzOKfJ2nm5moEF+iavkSlHGUp/p1JzTIt8zrE3FZre6CelfAwuH2kLRgGOxZZl/vsLBxgTabfQzh2un2x9c8DdeOdf6SJpWEApJIdbb9e3u3+5mj9sdt7rWvFVUG66dl59iIP5nG7mNpZ3uCzvYGHUvlTFZOka8L9vOkcagfjUL8aRzmR+NQfxqH+lOvpg/ubmYsFguLFh2hhq8uD0g1VLMBDFkEK1+Dla/bL7vFr4E+/8s7waMtB354FLbMsz++8RW48iHn1FxJKSCJXIw1GwzbxferTJIP24PRzh/tj4Mb2ddPa3RNiQ91Oj3bHn4cQch+SezAyXSsuau95mM2Qf1gX3v4CfOncag9CDUK9SfYT8u5iFyQmztc9ww0uhYWDLcP4J7ZE64fDx0fwmzLxm3BUPv6iCY3uG0atOnv7KorHQUkkaJYLbDsRdzXTOc2mwXjL0/7HR8efmf/9LX/6el77vuLPu973tfZ/T3P299czrei26z2S2lxL0D2GTC7Q9fHodsTF7ybxWozOHwq47yeoHOB6ERadpGv8/N0OxuAzoWgxmH+NKjli5e7brsX+VcadIGHVsH3j9unAlg6AbfdS7kyKQlz6vazi85+5JJzllUGCkgihUk5Cl8Nhfg15F6tN1mz7b1Jmcnld143z3wByvdswPIpJGQV9Xy+gJYbwlKOwI//B4c32M9Vr6P91v3wFo7Tp2XlsC/pbE/QsbNjhI6nsjcpjeyconvR6gR5FwhCjUL9CQ/00qSMIuXJp6Y9BP0xB356CvP+XwkFDE9/TP0/h4ZXO7vCSksBSSS/vcvt846kHQfPAHJumcqSXZn0uPYqPGzZYEkHS8bZP8//PgOy084+Pm+bJd+27HyvsaQDZy9FOULY6XJ7e4ZXICldn2Vr7TvZszedPWu3OoLQ+Uty5OfpbqZRiN+5EHQ2EDUM8dM6ZiLOZDJBu4FQ/0psXz9E5rE9eNz3BR71r3B2ZZWa/lUTyWWzwa9vwvJJ9jFH4ZdB308wAutj2bfIvsSGRzkMDDYM+6y2eQLUBUJYoSHrwiHMsKRjAGs8uzAu4z72LQoC1hdaTi0/z7OXws4NkG4c6k/dmj64aZV7EdcV0hTrkJ+I/fFHetdu4+xqKj0FJBGA9JOwcATsjrU/bnuffb0xDx+wWMr33CbTuUth/Pu7x863PymNN2P/4fsth3HHSk6m/a+8m9lEg2BfGhUIQn7U8NUgaZFKTZe1y4QCksihDfDFYEg5BO7e9hW1297n7Kr+lcSUTN6J28X89QfJsRmAid6t69O7VW2ahPlRP9gPT3dNFCciUhQFJKm+DAPWfQA/Pws2i/1W976f2Fe3rqSS0y3MWLmH2b/tI9NiH1R9XbNQnujVjJZ1tCCliEhxKSBJ9ZSZAt89Yr81FuDSW+G29yrtqtYZ2VY+Wr2f95fvJiXTPglj+wY1eapXMzo1quXk6kREKh8FJKl+Ev+GLwbBid32eYB6vgSdHqyU1+0tVhtfbDjI20t3cexMFgDNwgN4slczbrg0TLfYi4iUkgKSVC+b58EPMZCTYb8r7e6PILKjs6sqMZvN4Me/jvLmkp3sP5EOQL2aPsT0uITb2tTV3WYiIv+SApJUD5YM+Okp2PSJ/XHj6+HO/4Ff5br8ZBgGK3cl8driHfx9JAWw35b/yPVN6N+pvmanFhEpIwpIUvWd3Gu/pJbwF2Cyr2bd7YnyX9ajjG2KP8Vri3ewdu9JAPy93BnRrRH3X9UQf03UKCJSpvSvqlRt27+Hbx6GrBTwrWVf8brx9c6uqkR2JZ7h9Z93smRbIgCebmYGdW7Aw9c10cKuIiLlRAFJqiarBZZOhDXv2R9HdoK7ZkNQXaeWVRKHTqUzdekuFm46hM0Aswnual+Px7pfQt0aRS8sKyIi/54CklQ9KUfgy6FwcK39cedR0H0iuJXDMiHl4ERqFtN+2cOctQfIttrnMrqxZQRP9LqEJmEBTq5ORKR6UECSqmXPL/aFZtOTwCsQbpsGLW51dlXFkpqVw/9+3cuHK/eSlm0FoHOjWjx1YzPa1q/p5OpERKoXBSSpGmw2WPk6LJ8MGPbZsO/+GGo1dnZlF5WVY2Xe7/G8t2w3J9KyAbisbiBP9WrO1U1DNJeRiIgTKCBJ5Zd2AhYOhz1x9sftBsFNr51d/NV1WW0GX/9xmLdi/+Hw6QwAGob48X89L6H3ZbUxay4jERGnUUCSyu3gOvhyCKQcBncfuGUKtBng7KouyDAMYrcl8saSnfyTmApAeKAXj3e/hLva18PDTYvIiog4mwKSVE6GAb/PgCXjwJYDtZrYF5oNb+nsyi5o7d4TvLp4B3/EnwYgyMeDh69tzOAuUXh7VK55mUREqjIFJKl8MpPh21Gw/Tv745Z3wK3vgpfr3uH195FkXlu8kxX/HAfA28PM/V0b8p9rGhPkUznurhMRqU4UkKRySfjLPiv2yb1g9oBek6DjcJddaHZ/UhpTYv/huy1HAHA3m+jXMZJHr29KWKC3k6sTEZGiKCBJ5bHpU1j0BORkQlCkfaHZeh2cXVWhjqVk8s6yXXy+7iA5NgOAW1vXIabHJUSF+Dm5OhERuRinjwadNm0aUVFReHt706lTJ9atW1fkvhaLhRdeeIHGjRvj7e1N69atWbx4cYmPmZmZyciRI6lVqxb+/v706dOHxMTEMn9vUkay0+GbkfDdKHs4atID/rPSJcNRcoaF1xbvoNvrvzBnbTw5NoNrm4Xy46NX8U7/tgpHIiKVhFMD0vz584mJiWHChAls2rSJ1q1b06tXL44dO1bo/uPGjeO///0v7777Ltu2bePBBx/kjjvu4I8//ijRMUePHs3333/Pl19+yYoVKzhy5Ah33nlnub9fKYUTe2BmD9g8B0xmuH4cDPgCfIOdXVkeGdlWZqzYQ7fXfmH68j1kWmy0q1+Dz0dcyUdDO9KyTpCzSxQRkRJwakCaMmUKw4cPZ+jQobRo0YIZM2bg6+vLrFmzCt3/008/5ZlnnqF37940atSIhx56iN69e/Pmm28W+5jJycnMnDmTKVOmcP3119O+fXtmz57N6tWrWbt2bYW8bymmv7+B/14DiVvBLxQGfgPdngSz0zs+HSxWG/N+j+faN37hlZ92kJxh4ZJwfz4c1IEFD3Xhyka1nF2iiIiUgtPGIGVnZ7Nx40bGjh3r2GY2m+nevTtr1qwp9DVZWVl4e+cd2Orj48OqVauKfcyNGzdisVjo3r27Y5/mzZtTv3591qxZw5VXXlnkubOyshyPU1JSAPtlP4vFUpK3fkG5xyrLY1Y61mzMcc/jtv6/ANgir8R6x4cQUBuc8LkU1iY2m8HivxN5K243+0+kA1C3hjePXd+EW1vXxs1sIicnp8JrrQ70d8S1qD1ci9rj4or72TgtICUlJWG1WgkPD8+zPTw8nB07dhT6ml69ejFlyhS6detG48aNiYuLY+HChVit1mIfMyEhAU9PT2rUqFFgn4SEhCLrnTx5Ms8//3yB7UuWLMHX1/ei77ekYmNjy/yYlYF39gmu2Pcewel7ANgVdjPba92F8esfwB8XfnE5i42NxTBgZ7KJ7+PNHEqz3znn727Qs56NruGpuB/dzM9HNzu1zuqiuv4dcVVqD9ei9ihaenp6sfarVHexvf322wwfPpzmzZtjMplo3LgxQ4cOLfKSXFkaO3YsMTExjscpKSlERkbSs2dPAgMDy+w8FouF2NhYevTogYdH9Zofx7RnGW7fvogp4ySGVyDWW6cRdclNRDm5rtw2Cbu0I2//so+1+04B4OflxrCuUQzt0gB/r0r1V6lSq85/R1yR2sO1qD0uLvcK0MU47V/1kJAQ3NzcCtw9lpiYSERERKGvCQ0N5ZtvviEzM5MTJ05Qp04dxowZQ6NGjYp9zIiICLKzszl9+nSeXqQLnRfAy8sLLy+vAts9PDzK5YewvI7rkmxWWPEqrHgNMKB2a0x3f4x7cENnVwbArmOpzNxp5s81mwDwdDMzsHMDHr62MbX8C/5MSMWoVn9HKgG1h2tRexStuJ+L00a7enp60r59e+Li4hzbbDYbcXFxdO7c+YKv9fb2pm7duuTk5LBgwQJuu+22Yh+zffv2eHh45Nln586dxMfHX/S8Ug5Sj8OcO+0BCQPaD4X7l4ALhCPDMHhv2S5ueW81f540YzbB3e3r8cuT1zL+lhYKRyIiVZhTrwvExMQwePBgOnToQMeOHZk6dSppaWkMHToUgEGDBlG3bl0mT54MwO+//87hw4dp06YNhw8fZuLEidhsNp566qliHzMoKIhhw4YRExNDcHAwgYGBPPLII3Tu3LnIAdpSTuLX2heaPXMUPHzhlqnQ+h5nVwVApsXK0wv+5NvN9hmwW9W08ep9V9Gibk0nVyYiIhXBqQHpnnvu4fjx4zz33HMkJCTQpk0bFi9e7BhkHR8fj/m8W7ozMzMZN24ce/fuxd/fn969e/Ppp5/muVR2sWMCvPXWW5jNZvr06UNWVha9evVi+vTpFfa+qz3DgDXvQewEMKwQcol9odmwS51dGQDHzmQy4pONbD54GneziQm3XErg8T9pGubv7NJERKSCOH1k6ahRoxg1alShzy1fvjzP42uuuYZt27b9q2OC/RLdtGnTmDZtWolqlTKQcRq+HQk7frA/vqwPRL8DXq4RPrYeTmb4Jxs4mpxJDV8Ppt/bjivqB7Fo0Z/OLk1ERCqQ0wOSVCNH/7QvNHtqn32h2RsnwxUPuMxCs4u3HmX0/C1kWKw0DvVj5uAriArx03wiIiLVkAKSlD/DgE2fwKInwZoFQfWh70dQt72zKwPsg7Gn/bKbN5b8A0C3S0J5t39bgnx0B4iISHWlgCTlKzsdfvw/2DLP/viSG+H2911mLbVMi5UxC/7km7ODsYd0iWLczZfi7uY6y5mIiEjFU0CS8pO0y35J7di2swvNjoeuj7vMWmr5B2M/f1tL7u3UwNlliYiIC1BAkvKxK9Z+C392KviFwV2zoOHVzq7K4e8jyQz/eANHkjMJ8vHg/fva0aVxiLPLEhERF6GAJGVv/28w/z7IyYQGV8FdMyGg6FnKK9rirQmMnr+ZDIuVRmcHYzcM8XN2WSIi4kIUkKRsHd0Cn/Wzh6NLboJ7PgU31xjsbBgG05fv4fWfdwJwddMQ3hvQToOxRUSkAAUkKTtJu+HTOyErBRp0hbtnu0w4yj8Ye3DnBoy/pYUGY4uISKEUkKRsJB+GT2+H9CSIuBz6fwYePs6uCrAPxv7Ppxv5I/40bmYTz9/akvuu1GBsEREpmgKS/HtpJ+DTOyD5INRqAvctBO8gZ1cFwLYjKTzw8XrHYOzp97ajaxMNxhYRkQtTQJJ/J+sMzL0LknZCYF0Y+DX4hzq7KgB+/ts+GDs920qjED9mDtFgbBERKR4FJCm9nCz4/F44sgl8gu3hqEZ9Z1eFYRi8v8I+GNsw4KomIUwb0I4gX9cYDyUiIq5PAUlKx5oDC4bBvhXg6Q/3LYDQZs6uikyLlWcW/sXCPw4DGowtIiKlo4AkJWcY8MNjsP17cPOEfvOgbjtnV8XxM1n859MNbDo7GHvirS0ZqMHYIiJSCgpIUjKGAbHj4Y859uVD7poNja5xdlVsO5LC8E82cPh0BoHe7ky/tz1XNdVgbBERKR0FJCmZVW/B6nft39/6Llx6i3PrAZb8ncDj5w3G/t/gDjQK9Xd2WSIiUokpIEnxbZgNcc/bv+/5ErS9z6nlGIbBjBV7ee3nHRqMLSIiZUoBSYrn76/hh9H276+KgS6POLWcrBwrYxf+xcJN9sHYA69swHPRLfDQYGwRESkDCkhycbvjYMFwwID2Q+GG55xaTlJqFv/5dCMbD5zCzWxiQnQLBnWOcmpNIiJStSggyYUdXA/z7wObBVreATe/CSaT08rZfjSFBz4+Nxh72r3tuLqpa0xMKSIiVYcCkhQtcZt9lmxLOjS+Hu74AMxuTisndlsij33+B+nZVhqeHYzdWIOxRUSkHCggSeFO7bevr5Z5GupdAffMAXdPp5RiGAb/XbmXVxfbB2N3bVKLaQPaUcPXOfWIiEjVp4AkBZ1JhE9uh9QECGsBA74AT+esYZZ/MPZ9V9ZnQnRLDcYWEZFypYAkeWWchjl94NQ+qNEA7lsIvsFOKSUpNYsHP93IBg3GFhGRCqaAJOdkp8O8eyDxL/ALsy8+G1jbKaWcPxg7wNud6RqMLSIiFUgBSeysFvhyMBxcC15B9nBUq7FTSll6djB2WraVqFq+/G/wFTQJ02BsERGpOApIAjYbfPMQ7FoC7j5w7xcQcVmFl5F/MHaXxrWYfq8GY4uISMVTQKruDAN+egr++hLM7nDPp1D/ygovIyvHyjMLt7Jg0yEA7u1Un4m3ajC2iIg4hwJSdbd8Mqz/EDDBHf+Fpj0qvITzB2ObTTAhuiWDOjfA5MQJKUVEpHpTQKrO1s6AFa/av+/9OrS6q8JL2JGQwrCPzg3GnjagHd0u0WBsERFxLgWk6mrL57D4afv31z0LHYdXeAlx2xN59DP7YOwGtXyZqcHYIiLiIhSQqqOdP8E3D9u/7/QQdHuyQk9vGAYf/rqXyT/ZB2Nf2SiY9+9tT00/DcYWERHXoIBU3ez/Db4cAoYVLu8HvSZV6OKzWTlWxn29lS832gdj9+9Ynxdu02BsERFxLQpI1cnRLfBZP8jJhEtugtveA3PFBZMTqVk8OGcj6/fbB2OPv6UFQ7pEaTC2iIi4HAWk6iJpN3x6J2SlQIOucPdscPOosNPvTDjDsI/Xc+hUBgFe7rw7oC3XNgursPOLiIiUhAJSdZB8GD69HdKTIOJy6P8ZePhU2OmX7UjkkXnnD8buQJOwgAo7v4iISEkpIFV1aSfg0zsg+SDUamJffNY7qEJObRgG//t1H5N+2q7B2CIiUqkoIFVlWWdg7l2QtBMC69rXV/OvmDmGsnNsPPv1X+cNxo7k+Vsvw9Ndg7FFRMT1KSBVVTlZ8Pm9cGQT+ATbw1GN+hV2+neX7eLLjYcwm2DczS0Y2lWDsUVEpPJw+n/np02bRlRUFN7e3nTq1Il169ZdcP+pU6fSrFkzfHx8iIyMZPTo0WRmZjqej4qy/yLO/zVy5EjHPtdee22B5x988MFye48VzpoDC4bBvhXg6Q/3fQWhzSq0hLjtxwB44bbLuP+qhgpHIiJSqTi1B2n+/PnExMQwY8YMOnXqxNSpU+nVqxc7d+4kLKzgHU7z5s1jzJgxzJo1iy5duvDPP/8wZMgQTCYTU6ZMAWD9+vVYrVbHa7Zu3UqPHj24++678xxr+PDhvPDCC47Hvr6+5fQuK5hhwA+Pw/bvwc0T+s2Fuu0rtISUTAs7ElIA6NEivELPLSIiUhacGpCmTJnC8OHDGTp0KAAzZszgxx9/ZNasWYwZM6bA/qtXr6Zr164MGDAAsPcW9e/fn99//92xT2ho3jE2r7zyCo0bN+aaa67Js93X15eIiIiyfkvOF/sc/PEpmMxw1yxodG2Fl/BH/GlsBtQP9iU80LvCzy8iIvJvOS0gZWdns3HjRsaOHevYZjab6d69O2vWrCn0NV26dGHOnDmsW7eOjh07snfvXhYtWsTAgQOLPMecOXOIiYkpcIln7ty5zJkzh4iICKKjoxk/fvwFe5GysrLIyspyPE5JsfeQWCwWLBZLsd/3xeQeqzTHNK9+B7fV7wCQ0/stjCY3QhnWVly/70kCoH39oDL9bJzl37SJlD21h2tRe7gWtcfFFfezcVpASkpKwmq1Eh6e9xJMeHg4O3bsKPQ1AwYMICkpiauuugrDMMjJyeHBBx/kmWeeKXT/b775htOnTzNkyJACx2nQoAF16tThzz//5Omnn2bnzp0sXLiwyHonT57M888/X2D7kiVLyuXyXGxsbIn2b5D0C20OzgZga51+7DlSE44sKvO6imPJ32bAjGfKIRYtOuiUGspDSdtEypfaw7WoPVyL2qNo6enpxdrPZBiGUc61FOrIkSPUrVuX1atX07lzZ8f2p556ihUrVuS5bJZr+fLl9OvXj5deeolOnTqxe/duHnvsMYYPH8748eML7N+rVy88PT35/vvvL1jLsmXLuOGGG9i9ezeNGzcudJ/CepAiIyNJSkoiMDCwuG/7oiwWC7GxsfTo0QMPj+LNdG3a/h1uXz+AybBh7fIYtusKfhYVJTvHRvtJy8i02PjpkS40CfN3Wi1lpTRtIuVH7eFa1B6uRe1xcSkpKYSEhJCcnHzB399O60EKCQnBzc2NxMTEPNsTExOLHBs0fvx4Bg4cyAMPPABAq1atSEtLY8SIETz77LOYz1tX7MCBAyxduvSCvUK5OnXqBHDBgOTl5YWXl1eB7R4eHuXyQ1js4+5ZBt/8BwwbtB+CW4/ncXPiHWNbj54i02Kjpq8HzevUqFJ3r5VXW0vpqD1ci9rDtag9ilbcz8Vpt/l7enrSvn174uLiHNtsNhtxcXF5epTOl56enicEAbi5uQH2WZvPN3v2bMLCwrj55psvWsvmzZsBqF27dknegvMdXA+f3wc2C7S4HW6eAk4OJBv2nwKgfYPgKhWORESkenHqXWwxMTEMHjyYDh060LFjR6ZOnUpaWprjrrZBgwZRt25dJk+eDEB0dDRTpkyhbdu2jkts48ePJzo62hGUwB60Zs+ezeDBg3F3z/sW9+zZw7x58+jduze1atXizz//ZPTo0XTr1o3LL7+84t78v5W4zT5LtiUNGl0Hd34AZreLv66cbThwEoAOUTWdXImIiEjpOTUg3XPPPRw/fpznnnuOhIQE2rRpw+LFix0Dt+Pj4/P0GI0bNw6TycS4ceM4fPgwoaGhREdH8/LLL+c57tKlS4mPj+f+++8vcE5PT0+WLl3qCGORkZH06dOHcePGle+bLUun9tvXV8s8DfWugHvmgHvBy38VzTAMRw/SFQpIIiJSiTl9qZFRo0YxatSoQp9bvnx5nsfu7u5MmDCBCRMmXPCYPXv2LHDJLVdkZCQrVqwoVa0u4UwifHI7pCZAWAsY8AV4ucZA6H1JaZxIy8bT3cxldStmQVwREZHy4PSlRqQEMk7DnD5wah/UaAD3LQTfYGdX5ZDbe9SmXg283J1/uU9ERKS0FJAqi+x0mHcPJP4FfmH2xWcDXWtQ+fr9Gn8kIiJVgwJSZWC1wJeD4eBa8Aqyh6NahU9H4EwbDuSOP3KdXi0REZHSUEBydTYbfPMQ7FoC7j5w7xcQcZmzqyrg+Jks9iWlYTJBu/rqQRIRkcpNAcmVGQb89BT89SWY3eGeT6H+lc6uqlAbz97e3yw8gCBfTU4mIiKVmwKSK1s+GdZ/CJjgjv9C0x7OrqhI688O0Nb4IxERqQoUkFyUef0HsOJV+4Per0Oru5xb0EVsODtAW+OPRESkKnD6PEhSUL2Tv+H2x3/tD657FjoOd25BF5GencPWIykAtG+gHiQREan81IPkYkz/LKbtgQ/tDzo9BN2edG5BxbD54GmsNoPaQd7UreHj7HJERET+NQUkV5KTjduSZzBjw9aqL/Sa5PTFZ4tjg2P8kRaoFRGRqkEByZW4e5Iz4Cv2hVyP9ea3wVw5mme9Y/yRLq+JiEjVUDl+A1cnwY34M3IIuFWOW+VzrDY2nZ0gskMDDdAWEZGqQQFJ/pUdCWdIy7YS4OVOs4gAZ5cjIiJSJhSQ5F/Jvb2/XYOauJk1/khERKoGBST5V9Y71l/T+CMREak6FJCk1AzDcPQgddAEkSIiUoUoIEmpHTqVQWJKFh5uJlrXq+HsckRERMqMApKUWu7t/S3rBOHj6ebkakRERMqOApKU2gaNPxIRkSpKAUlKTeOPRESkqlJAklI5nZ7NP4mpAHTQArUiIlLFKCBJqWw8e3mtUagftfy9nFyNiIhI2VJAklJZf3aB2iu0vIiIiFRBCkhSKufGH+nymoiIVD0KSFJimRYrfx5KBuAKDdAWEZEqqMQBafbs2Xz55ZcFtn/55Zd8/PHHZVKUuLa/DieTbbUR4u9Fg1q+zi5HRESkzJU4IE2ePJmQkJAC28PCwpg0aVKZFCWuLXeCyCuiamIyaYFaERGpekockOLj42nYsGGB7Q0aNCA+Pr5MihLXtuHsAG3NfyQiIlVViQNSWFgYf/75Z4HtW7ZsoVatWmVSlLgum81w3OKv+Y9ERKSqKnFA6t+/P48++ii//PILVqsVq9XKsmXLeOyxx+jXr1951CguZPfxVJIzLPh4uNGiTqCzyxERESkX7iV9wYsvvsj+/fu54YYbcHe3v9xmszFo0CCNQaoGcscfta1fAw833QQpIiJVU4kDkqenJ/Pnz+ell15i8+bN+Pj40KpVKxo0aFAe9YmL0fgjERGpDkockHI1bdqUpk2blmUtUgmcfwebiIhIVVXiayR9+vTh1VdfLbD9tdde4+677y6TosQ1HU3O4NCpDMwmaFtfAUlERKquEgeklStX0rt37wLbb7rpJlauXFkmRYlryr281qJOIP5epe58FBERcXklDkipqal4enoW2O7h4UFKSkqZFCWuybH+mhaoFRGRKq7EAalVq1bMnz+/wPbPP/+cFi1alElR4prWn+1B0vprIiJS1ZX4Osn48eO588472bNnD9dffz0AcXFxzJs3j6+++qrMCxTXkJJpYUeCvYewgwZoi4hIFVfigBQdHc0333zDpEmT+Oqrr/Dx8aF169YsW7aM4GD1LFRVf8SfxmZAZLAP4YHezi5HRESkXJVqpr+bb76Z3377jbS0NPbu3Uvfvn154oknaN26dYmPNW3aNKKiovD29qZTp06sW7fugvtPnTqVZs2a4ePjQ2RkJKNHjyYzM9Px/MSJEzGZTHm+mjdvnucYmZmZjBw5klq1auHv70+fPn1ITEwsce3Vycbc2/s1/khERKqBUk+FvHLlSgYPHkydOnV48803uf7661m7dm2JjjF//nxiYmKYMGECmzZtonXr1vTq1Ytjx44Vuv+8efMYM2YMEyZMYPv27cycOZP58+fzzDPP5NmvZcuWHD161PG1atWqPM+PHj2a77//ni+//JIVK1Zw5MgR7rzzzpJ9ANXMek0QKSIi1UiJLrElJCTw0UcfMXPmTFJSUujbty9ZWVl88803pRqgPWXKFIYPH87QoUMBmDFjBj/++COzZs1izJgxBfZfvXo1Xbt2ZcCAAQBERUXRv39/fv/997xvyt2diIiIQs+ZnJzMzJkzmTdvnmMM1ezZs7n00ktZu3YtV155ZYnfR1Vnsdr442DuAG2NPxIRkaqv2AEpOjqalStXcvPNNzN16lRuvPFG3NzcmDFjRqlOnJ2dzcaNGxk7dqxjm9lspnv37qxZs6bQ13Tp0oU5c+awbt06OnbsyN69e1m0aBEDBw7Ms9+uXbuoU6cO3t7edO7cmcmTJ1O/fn0ANm7ciMVioXv37o79mzdvTv369VmzZk2RASkrK4usrCzH49wpDSwWCxaLpVSfQWFyj1WWx/y3thxKJtNio4aPB/VreLlUbRXBFdukOlN7uBa1h2tRe1xccT+bYgekn376iUcffZSHHnqoTJYYSUpKwmq1Eh4enmd7eHg4O3bsKPQ1AwYMICkpiauuugrDMMjJyeHBBx/Mc4mtU6dOfPTRRzRr1oyjR4/y/PPPc/XVV7N161YCAgJISEjA09OTGjVqFDhvQkJCkfVOnjyZ559/vsD2JUuW4OvrW4J3XjyxsbFlfszS+uWICXCjnncWixf/5OxynMaV2kTUHq5G7eFa1B5FS09PL9Z+xQ5Iq1atYubMmbRv355LL72UgQMH0q9fv1IXWBrLly9n0qRJTJ8+nU6dOrF7924ee+wxXnzxRcaPHw/YZ/TOdfnll9OpUycaNGjAF198wbBhw0p97rFjxxITE+N4nJKSQmRkJD179iQwMLD0byofi8VCbGwsPXr0wMPDo8yO+2/8MG8zcIybrmhG76sbOrucCueKbVKdqT1ci9rDtag9Lq64k1oXOyBdeeWVXHnllUydOpX58+cza9YsYmJisNlsxMbGEhkZSUBAQLELDAkJwc3NrcDdY4mJiUWOHxo/fjwDBw7kgQceAOyTVqalpTFixAieffZZzOaCY85r1KjBJZdcwu7duwGIiIggOzub06dP5+lFutB5Aby8vPDy8iqw3cPDo1x+CMvruCVlGAab4k8DcGXjEJeoyVlcpU3ETu3hWtQerkXtUbTifi4lvovNz8+P+++/n1WrVvHXX3/xf//3f7zyyiuEhYVx6623Fvs4np6etG/fnri4OMc2m81GXFwcnTt3LvQ16enpBUKQm5sbYP9FXpjU1FT27NlD7dq1AWjfvj0eHh55zrtz507i4+OLPG91ti8pjRNp2Xi6m7msbpCzyxEREakQpb7NH6BZs2a89tprHDp0iM8++6zEr4+JieHDDz/k448/Zvv27Tz00EOkpaU57mobNGhQnkHc0dHRvP/++3z++efs27eP2NhYxo8fT3R0tCMoPfHEE6xYsYL9+/ezevVq7rjjDtzc3Ojfvz8AQUFBDBs2jJiYGH755Rc2btzI0KFD6dy5s+5gK0TuArVt6tXAy93NydWIiIhUjDJZkt3NzY3bb7+d22+/vUSvu+eeezh+/DjPPfccCQkJtGnThsWLFzsGbsfHx+fpMRo3bhwmk4lx48Zx+PBhQkNDiY6O5uWXX3bsc+jQIfr378+JEycIDQ3lqquuYu3atYSGhjr2eeuttzCbzfTp04esrCx69erF9OnT/92HUEWtPztBZHvd3i8iItVImQSkf2PUqFGMGjWq0OeWL1+e57G7uzsTJkxgwoQJRR7v888/v+g5vb29mTZtGtOmTStRrdXRxgOa/0hERKqff3WJTaq2pNQs9ialAdC+vmbQFhGR6kMBSYqUO/6oWXgAQb66G0JERKoPBSQp0oaz44866PKaiIhUMwpIUqT1jvFHurwmIiLViwKSFCo9O4e/DycD6kESEZHqRwFJCrX54GlybAa1g7ypW8PH2eWIiIhUKAUkKVTuAO0OUcGYTCYnVyMiIlKxFJCkULkTRGr+IxERqY4UkKQAq83gj7ML1LZvoIAkIiLVjwKSFLAjIYXUrBz8vdxpHhHo7HJEREQqnAKSFJA7/qhdg5q4mTX+SEREqh8FJCnAMf5Il9dERKSaUkCSPAzDcASkDpogUkREqikFJMnj0KkMElOycDebaBNZw9nliIiIOIUCkuSx4YC99+iyukH4eLo5uRoRERHnUECSPNbvz11/TeOPRESk+lJAkjw2aPyRiIiIApKcczo9m38SUwHooDvYRESkGlNAEoeNB+yX1xqF+FHL38vJ1YiIiDiPApI4rHcsUKveIxERqd4UkMRh4wGNPxIREQEFJDkr02Jly8FkAK5QQBIRkWpOAUkA2Ho4mWyrjRB/T6Jq+Tq7HBEREadSQBLgvPFHDYIxmbRArYiIVG8KSAKcP/+RBmiLiIgoIAk2m8GGA7kzaGv8kYiIiAKSsPt4KskZFnw83GhRJ9DZ5YiIiDidApKw/uzltbb1a+Dhph8JERER/TYUNjgGaGv8kYiICCggCed6kDRBpIiIiJ0CUjWXkJzJoVMZmE32S2wiIiKigFTtbTi7vMiltQMJ8PZwcjUiIiKuQQGpmssdf6Tb+0VERM5RQKrm1muCSBERkQIUkKqxM5kWth9NAexLjIiIiIidAlI19kf8aWwGRAb7EBHk7exyREREXIYCUjWWu/7aFeo9EhERyUMBqRpbnztBpAZoi4iI5KGAVE1ZrDb+OJh7B5sGaIuIiJzP6QFp2rRpREVF4e3tTadOnVi3bt0F9586dSrNmjXDx8eHyMhIRo8eTWZmpuP5yZMnc8UVVxAQEEBYWBi33347O3fuzHOMa6+9FpPJlOfrwQcfLJf356r+PpJCpsVGkI8HjUP9nV2OiIiIS3FqQJo/fz4xMTFMmDCBTZs20bp1a3r16sWxY8cK3X/evHmMGTOGCRMmsH37dmbOnMn8+fN55plnHPusWLGCkSNHsnbtWmJjY7FYLPTs2ZO0tLQ8xxo+fDhHjx51fL322mvl+l5dTe74ow4NamI2m5xcjYiIiGtxd+bJp0yZwvDhwxk6dCgAM2bM4Mcff2TWrFmMGTOmwP6rV6+ma9euDBgwAICoqCj69+/P77//7thn8eLFeV7z0UcfERYWxsaNG+nWrZtju6+vLxEREeXxtiqFDRp/JCIiUiSnBaTs7Gw2btzI2LFjHdvMZjPdu3dnzZo1hb6mS5cuzJkzh3Xr1tGxY0f27t3LokWLGDhwYJHnSU5OBiA4OG8QmDt3LnPmzCEiIoLo6GjGjx+Pr69vkcfJysoiKyvL8TglxT5/kMViwWKxXPwNF1PuscrymPkZhuGYILJtvYByPVdVUBFtIsWn9nAtag/Xova4uOJ+Nk4LSElJSVitVsLDw/NsDw8PZ8eOHYW+ZsCAASQlJXHVVVdhGAY5OTk8+OCDeS6xnc9ms/H444/TtWtXLrvssjzHadCgAXXq1OHPP//k6aefZufOnSxcuLDIeidPnszzzz9fYPuSJUsuGKxKKzY2tsyPmetYBpxIc8fdZHD4rzUk/l1up6pSyrNNpOTUHq5F7eFa1B5FS09PL9Z+Tr3EVlLLly9n0qRJTJ8+nU6dOrF7924ee+wxXnzxRcaPH19g/5EjR7J161ZWrVqVZ/uIESMc37dq1YratWtzww03sGfPHho3blzouceOHUtMTIzjcUpKCpGRkfTs2ZPAwMAyeof2ZBsbG0uPHj3w8CifxWO/2nQYNv9Nm/o1ufWWjuVyjqqkItpEik/t4VrUHq5F7XFxuVeALsZpASkkJAQ3NzcSExPzbE9MTCxybND48eMZOHAgDzzwAGAPN2lpaYwYMYJnn30Ws/ncmPNRo0bxww8/sHLlSurVq3fBWjp16gTA7t27iwxIXl5eeHl5Fdju4eFRLj+E5XVcgD8O2i87XtGwlv4ClUB5tomUnNrDtag9XIvao2jF/Vycdhebp6cn7du3Jy4uzrHNZrMRFxdH586dC31Nenp6nhAE4ObmBtjH1eT+OWrUKL7++muWLVtGw4YNL1rL5s2bAahdu3Zp3kqlkztAW/MfiYiIFM6pl9hiYmIYPHgwHTp0oGPHjkydOpW0tDTHXW2DBg2ibt26TJ48GYDo6GimTJlC27ZtHZfYxo8fT3R0tCMojRw5knnz5vHtt98SEBBAQkICAEFBQfj4+LBnzx7mzZtH7969qVWrFn/++SejR4+mW7duXH755c75ICpQUmoWe5PsUx60r6872ERERArj1IB0zz33cPz4cZ577jkSEhJo06YNixcvdgzcjo+Pz9NjNG7cOEwmE+PGjePw4cOEhoYSHR3Nyy+/7Njn/fffB+yTQZ5v9uzZDBkyBE9PT5YuXeoIY5GRkfTp04dx48aV/xt2Abm9R83CAwjyVferiIhIYZw+SHvUqFGMGjWq0OeWL1+e57G7uzsTJkxgwoQJRR4v91JbUSIjI1mxYkWJ66wqcieIbK/LayIiIkVy+lIjUrHWH9D4IxERkYtRQKpGMrKt/H3YfgdbhwYafyQiIlIUBaRqZPPB0+TYDCICvalX08fZ5YiIiLgsBaRqxLFAbVRNTCYtUCsiIlIUBaRq5Nz4I11eExERuRAFpGrCajPYdDYgddAAbRERkQtSQKomdiSkkJqVg7+XO80jym7tOBERkapIAamayJ0gsl2DmriZNf5IRETkQhSQqon1ZwdoX9FAl9dEREQuRgGpGjAMwxGQNIO2iIjIxSkgVQOHTmWQmJKFu9lEm8gazi5HRETE5SkgVQMbz9691rJuEL6eTl9+T0RExOUpIFUDGn8kIiJSMgpI1UDuHWwdNEGkiIhIsSggVXHJ6RZ2Jp4BNEGkiIhIcSkgVXEb4+2X1xqF+BHi7+XkakRERCoHBaQqbv1+LS8iIiJSUgpIVdyGswO0Nf5IRESk+BSQqrBMi5UtB5MBuEIBSUREpNgUkKqwrYeTybbaqOXnSVQtX2eXIyIiUmkoIFVh548/Mpm0QK2IiEhxKSBVYbnjj3R5TUREpGQUkKoom81gY7wmiBQRESkNBaQqas/xVE6nW/D2MNOyTqCzyxEREalUFJCqqNzxR20ja+LhpmYWEREpCf3mrKLOjT/SBJEiIiIlpYBURa0/oAkiRURESksBqQpKSM7k4MkMzCZoW7+Gs8sRERGpdBSQqqANZ3uPLq0dSIC3h5OrERERqXwUkKqgDbkTRDbQ+CMREZHSUECqgtZrgVoREZF/RQGpiknNymH70RTAvsSIiIiIlJwCUhXzR/wpbAbUq+lD7SAfZ5cjIiJSKSkgVTG5E0Rq/TUREZHSU0CqYjY4xh/p8pqIiEhpKSBVIRarjT/iTwPqQRIREfk3FJCqkG1HUsiwWAny8aBJqL+zyxEREam0FJCqEMft/Q1qYjabnFyNiIhI5aWAVIU4JojU5TUREZF/xekBadq0aURFReHt7U2nTp1Yt27dBfefOnUqzZo1w8fHh8jISEaPHk1mZmaJjpmZmcnIkSOpVasW/v7+9OnTh8TExDJ/bxXJMAzHEiMaoC0iIvLvODUgzZ8/n5iYGCZMmMCmTZto3bo1vXr14tixY4XuP2/ePMaMGcOECRPYvn07M2fOZP78+TzzzDMlOubo0aP5/vvv+fLLL1mxYgVHjhzhzjvvLPf3W572n0gnKTUbTzczreoGObscERGRSs2pAWnKlCkMHz6coUOH0qJFC2bMmIGvry+zZs0qdP/Vq1fTtWtXBgwYQFRUFD179qR///55eogudszk5GRmzpzJlClTuP7662nfvj2zZ89m9erVrF27tkLed3nIHX90eb0gvD3cnFyNiIhI5eburBNnZ2ezceNGxo4d69hmNpvp3r07a9asKfQ1Xbp0Yc6cOaxbt46OHTuyd+9eFi1axMCBA4t9zI0bN2KxWOjevbtjn+bNm1O/fn3WrFnDlVdeWei5s7KyyMrKcjxOSbEv52GxWLBYLKX8FArKPVZJj7l+3wkA2tUPKtN6pPRtIuVD7eFa1B6uRe1xccX9bJwWkJKSkrBarYSHh+fZHh4ezo4dOwp9zYABA0hKSuKqq67CMAxycnJ48MEHHZfYinPMhIQEPD09qVGjRoF9EhISiqx38uTJPP/88wW2L1myBF9f34u+35KKjY0t0f4r/nYDTBjH97Bo0e4yr0dK3iZSvtQerkXt4VrUHkVLT08v1n5OC0ilsXz5ciZNmsT06dPp1KkTu3fv5rHHHuPFF19k/Pjx5XrusWPHEhMT43ickpJCZGQkPXv2JDAwsMzOY7FYiI2NpUePHnh4eBTrNSfSsjm2ZjkAI+7oTg3f4r1Oiqc0bSLlR+3hWtQerkXtcXG5V4AuxmkBKSQkBDc3twJ3jyUmJhIREVHoa8aPH8/AgQN54IEHAGjVqhVpaWmMGDGCZ599tljHjIiIIDs7m9OnT+fpRbrQeQG8vLzw8vIqsN3Dw6NcfghLctwth+2X1y4J9yc0qOx7s8SuvNpaSkft4VrUHq5F7VG04n4uThuk7enpSfv27YmLi3Nss9lsxMXF0blz50Jfk56ejtmct2Q3N/uAZMMwinXM9u3b4+HhkWefnTt3Eh8fX+R5Xd259dc0/5GIiEhZcOoltpiYGAYPHkyHDh3o2LEjU6dOJS0tjaFDhwIwaNAg6taty+TJkwGIjo5mypQptG3b1nGJbfz48URHRzuC0sWOGRQUxLBhw4iJiSE4OJjAwEAeeeQROnfuXOQAbVe3/uwEkVdo/iMREZEy4dSAdM8993D8+HGee+45EhISaNOmDYsXL3YMso6Pj8/TYzRu3DhMJhPjxo3j8OHDhIaGEh0dzcsvv1zsYwK89dZbmM1m+vTpQ1ZWFr169WL69OkV98bLUEa2la2HkwHo0EA9SCIiImXB6YO0R40axahRowp9bvny5Xkeu7u7M2HCBCZMmFDqYwJ4e3szbdo0pk2bVuJ6Xc3mg6fJsRlEBHpTr6aPs8sRERGpEpy+1Ij8O7njj9pH1cRk0gK1IiIiZUEBqZJbf+Ds+KMGGn8kIiJSVhSQKjGrzWDT2YCkO9hERETKjgJSJbYz4QypWTn4e7nTPCLA2eWIiIhUGQpIldiGA/bxR23r18DdTU0pIiJSVvRbtRI7N/+RLq+JiIiUJQWkSsowDNbvy51BWwO0RUREypICUiV1+HQGCSmZuJtNtIms4exyREREqhQFpEpqw9nLay3rBuHr6fT5PkVERKoUBaRKav3ZCSI1/5GIiEjZU0CqpHJ7kDT+SEREpOwpIFVCyekWdiaeAaC9FqgVEREpcwpIldDGePvltYYhfoQGeDm5GhERkapHAakSclxe0/gjERGRcqGAVAlt0ASRIiIi5UoBqZLJyrGy+dBpQAO0RUREyosCUiWz9XAy2Tk2avl50jDEz9nliIiIVEkKSJXM+vNu7zeZTE6uRkREpGpSQKpkNuROEKnxRyIiIuVGAakSsdkMNhzI7UFSQBIRESkvCkiVyJ7jqZxOt+DtYaZlnUBnlyMiIlJlKSBVIrnjj9pE1sDDTU0nIiJSXvRbthLR+CMREZGKoYBUiaw/YA9IGn8kIiJSvhSQKonElEwOnszAbIJ29Ws4uxwREZEqTQGpkshdXqR5RCAB3h5OrkZERKRqU0CqJNY7xh9peREREZHypoBUSWzQ+CMREZEKo4BUCaRm5bDtSAqgBWpFREQqggJSJfBH/ClsBtSr6UPtIB9nlyMiIlLlKSBVAo4Fahuo90hERKQiKCBVArkTRGr8kYiISMVQQHJxFquNP+JPA5pBW0REpKIoILm4bUdSyLBYCfR2p2mYv7PLERERqRYUkFzchgNnxx9FBWM2m5xcjYiISPWggOTizo0/0gBtERGRiqKA5MIMw3DcwabxRyIiIhVHAcmFHTiRTlJqFp5uZlrVDXJ2OSIiItWGApILy11/7fJ6QXh7uDm5GhERkerDJQLStGnTiIqKwtvbm06dOrFu3boi97322msxmUwFvm6++WbHPoU9bzKZeP311x37REVFFXj+lVdeKdf3WVIb9p8boC0iIiIVx93ZBcyfP5+YmBhmzJhBp06dmDp1Kr169WLnzp2EhYUV2H/hwoVkZ2c7Hp84cYLWrVtz9913O7YdPXo0z2t++uknhg0bRp8+ffJsf+GFFxg+fLjjcUBAQFm9rTKxPneBWs2gLSIiUqGcHpCmTJnC8OHDGTp0KAAzZszgxx9/ZNasWYwZM6bA/sHBeXtTPv/8c3x9ffMEpIiIiDz7fPvtt1x33XU0atQoz/aAgIAC+7qKE2nZ7D2eBkB7BSQREZEK5dSAlJ2dzcaNGxk7dqxjm9lspnv37qxZs6ZYx5g5cyb9+vXDz8+v0OcTExP58ccf+fjjjws898orr/Diiy9Sv359BgwYwOjRo3F3L/wjycrKIisry/E4JSUFAIvFgsViKVatxZF7rHV7kwBoEuqHv6epTM8hJZP72asNXIPaw7WoPVyL2uPiivvZODUgJSUlYbVaCQ8Pz7M9PDycHTt2XPT169atY+vWrcycObPIfT7++GMCAgK4884782x/9NFHadeuHcHBwaxevZqxY8dy9OhRpkyZUuhxJk+ezPPPP19g+5IlS/D19b1orSX19a9/AmbCTGdYtGhRmR9fSi42NtbZJch51B6uRe3hWtQeRUtPTy/Wfk6/xPZvzJw5k1atWtGxY8ci95k1axb33nsv3t7eebbHxMQ4vr/88svx9PTkP//5D5MnT8bLy6vAccaOHZvnNSkpKURGRtKzZ08CAwPL4N3YWSwWYmNjOWmuAaRwx9WX07tNnTI7vpRcbpv06NEDDw8PZ5dT7ak9XIvaw7WoPS4u9wrQxTg1IIWEhODm5kZiYmKe7YmJiRcdG5SWlsbnn3/OCy+8UOQ+v/76Kzt37mT+/PkXraVTp07k5OSwf/9+mjVrVuB5Ly+vQoOTh4dHmf8QZlthW8IZAK5sHKofchdRHm0tpaf2cC1qD9ei9ihacT8Xp97m7+npSfv27YmLi3Nss9lsxMXF0blz5wu+9ssvvyQrK4v77ruvyH1mzpxJ+/btad269UVr2bx5M2azudA75ypafCpYrAbhgV7Uq+nj7HJERESqHadfYouJiWHw4MF06NCBjh07MnXqVNLS0hx3tQ0aNIi6desyefLkPK+bOXMmt99+O7Vq1Sr0uCkpKXz55Ze8+eabBZ5bs2YNv//+O9dddx0BAQGsWbOG0aNHc99991GzpvPvGNt7xr4obYeoYEwmLVArIiJS0ZwekO655x6OHz/Oc889R0JCAm3atGHx4sWOgdvx8fGYzXk7unbu3MmqVatYsmRJkcf9/PPPMQyD/v37F3jOy8uLzz//nIkTJ5KVlUXDhg0ZPXp0njFGzpQbkK7Q7f0iIiJO4fSABDBq1ChGjRpV6HPLly8vsK1Zs2YYhnHBY44YMYIRI0YU+ly7du1Yu3ZtieusCFabwb7zepBERESk4rnEUiNyzj+JqWRaTfh5utE8wrVm9hYREakuFJBczMZ4+/prbSJr4O6m5hEREXEG/QZ2MRsOnAagfYMaTq1DRESkOlNAcjEZ2VZMGHRQQBIREXEaBSQX89/72vJqRysddAebiIiI0ygguSAvN/DQ+CMRERGn0W9hERERkXwUkERERETyUUASERERyUcBSURERCQfBSQRERGRfBSQRERERPJRQBIRERHJRwFJREREJB8FJBEREZF8FJBERERE8lFAEhEREclHAUlEREQkHwUkERERkXzcnV1AZWUYBgApKSllelyLxUJ6ejopKSl4eHiU6bGldNQmrkXt4VrUHq5F7XFxub+3c3+PF0UBqZTOnDkDQGRkpJMrERERkZI6c+YMQUFBRT5vMi4WoaRQNpuNI0eOEBAQgMlkKrPjpqSkEBkZycGDBwkMDCyz40rpqU1ci9rDtag9XIva4+IMw+DMmTPUqVMHs7nokUbqQSols9lMvXr1yu34gYGB+uF2MWoT16L2cC1qD9ei9riwC/Uc5dIgbREREZF8FJBERERE8lFAcjFeXl5MmDABLy8vZ5ciZ6lNXIvaw7WoPVyL2qPsaJC2iIiISD7qQRIRERHJRwFJREREJB8FJBEREZF8FJBERERE8lFAcjHTpk0jKioKb29vOnXqxLp165xdUrU0efJkrrjiCgICAggLC+P2229n586dzi5LznrllVcwmUw8/vjjzi6l2jp8+DD33XcftWrVwsfHh1atWrFhwwZnl1VtWa1Wxo8fT8OGDfHx8aFx48a8+OKLF11vTIqmgORC5s+fT0xMDBMmTGDTpk20bt2aXr16cezYMWeXVu2sWLGCkSNHsnbtWmJjY7FYLPTs2ZO0tDRnl1btrV+/nv/+979cfvnlzi6l2jp16hRdu3bFw8ODn376iW3btvHmm29Ss2ZNZ5dWbb366qu8//77vPfee2zfvp1XX32V1157jXfffdfZpVVaus3fhXTq1IkrrriC9957D7Cv9xYZGckjjzzCmDFjnFxd9Xb8+HHCwsJYsWIF3bp1c3Y51VZqairt2rVj+vTpvPTSS7Rp04apU6c6u6xqZ8yYMfz222/8+uuvzi5FzrrlllsIDw9n5syZjm19+vTBx8eHOXPmOLGyyks9SC4iOzubjRs30r17d8c2s9lM9+7dWbNmjRMrE4Dk5GQAgoODnVxJ9TZy5EhuvvnmPH9PpOJ99913dOjQgbvvvpuwsDDatm3Lhx9+6OyyqrUuXboQFxfHP//8A8CWLVtYtWoVN910k5Mrq7y0WK2LSEpKwmq1Eh4enmd7eHg4O3bscFJVAvaevMcff5yuXbty2WWXObucauvzzz9n06ZNrF+/3tmlVHt79+7l/fffJyYmhmeeeYb169fz6KOP4unpyeDBg51dXrU0ZswYUlJSaN68OW5ublitVl5++WXuvfdeZ5dWaSkgiVzEyJEj2bp1K6tWrXJ2KdXWwYMHeeyxx4iNjcXb29vZ5VR7NpuNDh06MGnSJADatm3L1q1bmTFjhgKSk3zxxRfMnTuXefPm0bJlSzZv3szjjz9OnTp11CalpIDkIkJCQnBzcyMxMTHP9sTERCIiIpxUlYwaNYoffviBlStXUq9ePWeXU21t3LiRY8eO0a5dO8c2q9XKypUree+998jKysLNzc2JFVYvtWvXpkWLFnm2XXrppSxYsMBJFcmTTz7JmDFj6NevHwCtWrXiwIEDTJ48WQGplDQGyUV4enrSvn174uLiHNtsNhtxcXF07tzZiZVVT4ZhMGrUKL7++muWLVtGw4YNnV1StXbDDTfw119/sXnzZsdXhw4duPfee9m8ebPCUQXr2rVrgWkv/vnnHxo0aOCkiiQ9PR2zOe+vdDc3N2w2m5MqqvzUg+RCYmJiGDx4MB06dKBjx45MnTqVtLQ0hg4d6uzSqp2RI0cyb948vv32WwICAkhISAAgKCgIHx8fJ1dX/QQEBBQY/+Xn50etWrU0LswJRo8eTZcuXZg0aRJ9+/Zl3bp1fPDBB3zwwQfOLq3aio6O5uWXX6Z+/fq0bNmSP/74gylTpnD//fc7u7RKS7f5u5j33nuP119/nYSEBNq0acM777xDp06dnF1WtWMymQrdPnv2bIYMGVKxxUihrr32Wt3m70Q//PADY8eOZdeuXTRs2JCYmBiGDx/u7LKqrTNnzjB+/Hi+/vprjh07Rp06dejfvz/PPfccnp6ezi6vUlJAEhEREclHY5BERERE8lFAEhEREclHAUlEREQkHwUkERERkXwUkERERETyUUASERERyUcBSURERCQfBSQRkVKKiorSRJUiVZQCkohUCkOGDOH2228H7LNoP/744xV27o8++ogaNWoU2L5+/XpGjBhRYXWISMXRWmwiUm1lZ2f/q2UYQkNDy7AaEXEl6kESkUplyJAhrFixgrfffhuTyYTJZGL//v0AbN26lZtuugl/f3/Cw8MZOHAgSUlJjtdee+21jBo1iscff5yQkBB69eoFwJQpU2jVqhV+fn5ERkby8MMPk5qaCsDy5csZOnQoycnJjvNNnDgRKHiJLT4+nttuuw1/f38CAwPp27cviYmJjucnTpxImzZt+PTTT4mKiiIoKIh+/fpx5syZ8v3QRKTEFJBEpFJ5++236dy5M8OHD+fo0aMcPXqUyMhITp8+zfXXX0/btm3ZsGEDixcvJjExkb59++Z5/ccff4ynpye//fYbM2bMAMBsNvPOO+/w999/8/HHH7Ns2TKeeuopALp06cLUqVMJDAx0nO+JJ54oUJfNZuO2227j5MmTrFixgtjYWPbu3cs999yTZ789e/bwzTff8MMPP/DDDz+wYsUKXnnllXL6tESktHSJTUQqlaCgIDw9PfH19SUiIsKx/b333qNt27ZMmjTJsW3WrFlERkbyzz//cMkllwDQtGlTXnvttTzHPH88U1RUFC+99BIPPvgg06dPx9PTk6CgIEwmU57z5RcXF8dff/3Fvn37iIyMBOCTTz6hZcuWrF+/niuuuAKwB6mPPvqIgIAAAAYOHEhcXBwvv/zyv/tgRKRMqQdJRKqELVu28Msvv+Dv7+/4at68OWDvtcnVvn37Aq9dunQpN9xwA3Xr1iUgIICBAwdy4sQJ0tPTi33+7du3ExkZ6QhHAC1atKBGjRps377dsS0qKsoRjgBq167NsWPHSvReRaT8qQdJRKqE1NRUoqOjefXVVws8V7t2bcf3fn5+eZ7bv38/t9xyCw899BAvv/wywcHBrFq1imHDhpGdnY2vr2+Z1unh4ZHnsclkwmazlek5ROTfU0ASkUrH09MTq9WaZ1u7du1YsGABUVFRuLsX/5+2jRs3YrPZePPNNzGb7Z3qX3zxxUXPl9+ll17KwYMHOXjwoKMXadu2bZw+fZoWLVoUux4RcQ26xCYilU5UVBS///47+/fvJykpCZvNxsiRIzl58iT9+/dn/fr17Nmzh59//pmhQ4deMNw0adIEi8XCu+++y969e/n0008dg7fPP19qaipxcXEkJSUVeumte/futGrVinvvvZdNmzaxbt06Bg0axDXXXEOHDh3K/DMQkfKlgCQilc4TTzyBm5sbLVq0IDQ0lPj4eOrUqcNvv/2G1WqlZ8+etGrViscff5waNWo4eoYK07p1a6ZMmcKrr77KZZddxty5c5k8eXKefbp06cKDDz7IPffcQ2hoaIFB3mC/VPbtt99Ss2ZNunXrRvfu3WnUqBHz588v8/cvIuXPZBiG4ewiRERERFyJepBERERE8lFAEhEREclHAUlEREQkHwUkERERkXwUkERERETyUUASERERyUcBSURERCQfBSQRERGRfBSQRERERPJRQBIRERHJRwFJREREJB8FJBEREZF8/h9ECIi0HXxsgAAAAABJRU5ErkJggg==",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# vizualize mnist\n",
    "\n",
    "for item in history.history:\n",
    "    print(\"Key:\",item)\n",
    "\n",
    "plt.plot(history.history['loss'],label=\"train\")\n",
    "plt.plot(history.history['val_loss'],label=\"validation\")\n",
    "plt.title('Model Loss')\n",
    "plt.yscale('log')\n",
    "plt.ylabel('Cross Entropy')\n",
    "plt.xlabel('Iteration')\n",
    "plt.grid()\n",
    "plt.legend()\n",
    "plt.show()\n",
    "\n",
    "#print(\"history\",history.history)\n",
    "plt.plot(history.history['accuracy'],label=\"train\")\n",
    "plt.plot(history.history['val_accuracy'],label=\"validation\")\n",
    "plt.title('Model Accuracy')\n",
    "#plt.yscale('log')\n",
    "plt.ylabel('Acc')\n",
    "plt.xlabel('Iteration')\n",
    "plt.grid()\n",
    "plt.legend()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 3.\n",
    "\n",
    "Realizuoti U-Net tinklą."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 100,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "WARNING:tensorflow:`input_shape` is undefined or non-square, or `rows` is not in [96, 128, 160, 192, 224]. Weights for input shape (224, 224) will be loaded as the default.\n",
      "Epoch 1/20\n",
      "101/938 [==>...........................] - ETA: 3:20 - loss: 12.1178 - accuracy: 0.5198"
     ]
    },
    {
     "ename": "KeyboardInterrupt",
     "evalue": "",
     "output_type": "error",
     "traceback": [
      "\u001b[1;31m---------------------------------------------------------------------------\u001b[0m",
      "\u001b[1;31mKeyboardInterrupt\u001b[0m                         Traceback (most recent call last)",
      "Cell \u001b[1;32mIn[100], line 69\u001b[0m\n\u001b[0;32m     61\u001b[0m model\u001b[38;5;241m.\u001b[39mcompile(\n\u001b[0;32m     62\u001b[0m   optimizer\u001b[38;5;241m=\u001b[39m\u001b[38;5;124m'\u001b[39m\u001b[38;5;124madam\u001b[39m\u001b[38;5;124m'\u001b[39m,\n\u001b[0;32m     63\u001b[0m   loss\u001b[38;5;241m=\u001b[39mtf\u001b[38;5;241m.\u001b[39mkeras\u001b[38;5;241m.\u001b[39mlosses\u001b[38;5;241m.\u001b[39mSparseCategoricalCrossentropy(from_logits\u001b[38;5;241m=\u001b[39m\u001b[38;5;28;01mFalse\u001b[39;00m),\n\u001b[0;32m     64\u001b[0m   metrics\u001b[38;5;241m=\u001b[39m[\u001b[38;5;124m'\u001b[39m\u001b[38;5;124maccuracy\u001b[39m\u001b[38;5;124m'\u001b[39m]\n\u001b[0;32m     65\u001b[0m )\n\u001b[0;32m     67\u001b[0m \u001b[38;5;66;03m#tf.keras.utils.plot_model(model, show_shapes=True)\u001b[39;00m\n\u001b[1;32m---> 69\u001b[0m model_history \u001b[38;5;241m=\u001b[39m \u001b[43mmodel\u001b[49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mfit\u001b[49m\u001b[43m(\u001b[49m\n\u001b[0;32m     70\u001b[0m \u001b[43m  \u001b[49m\u001b[43mtrain_images\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43mtrain_labels0\u001b[49m\u001b[43m,\u001b[49m\n\u001b[0;32m     71\u001b[0m \u001b[43m  \u001b[49m\u001b[43mepochs\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[38;5;241;43m20\u001b[39;49m\u001b[43m,\u001b[49m\n\u001b[0;32m     72\u001b[0m \u001b[43m  \u001b[49m\u001b[43mbatch_size\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[38;5;241;43m64\u001b[39;49m\u001b[43m,\u001b[49m\n\u001b[0;32m     73\u001b[0m \u001b[43m  \u001b[49m\u001b[43mvalidation_data\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[43m(\u001b[49m\u001b[43mtest_images\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43mtest_labels0\u001b[49m\u001b[43m)\u001b[49m\n\u001b[0;32m     74\u001b[0m \u001b[43m)\u001b[49m\n",
      "File \u001b[1;32mf:\\KTU\\Neuroninių tinklų metodai\\venv\\Lib\\site-packages\\keras\\src\\utils\\traceback_utils.py:65\u001b[0m, in \u001b[0;36mfilter_traceback.<locals>.error_handler\u001b[1;34m(*args, **kwargs)\u001b[0m\n\u001b[0;32m     63\u001b[0m filtered_tb \u001b[38;5;241m=\u001b[39m \u001b[38;5;28;01mNone\u001b[39;00m\n\u001b[0;32m     64\u001b[0m \u001b[38;5;28;01mtry\u001b[39;00m:\n\u001b[1;32m---> 65\u001b[0m     \u001b[38;5;28;01mreturn\u001b[39;00m \u001b[43mfn\u001b[49m\u001b[43m(\u001b[49m\u001b[38;5;241;43m*\u001b[39;49m\u001b[43margs\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[38;5;241;43m*\u001b[39;49m\u001b[38;5;241;43m*\u001b[39;49m\u001b[43mkwargs\u001b[49m\u001b[43m)\u001b[49m\n\u001b[0;32m     66\u001b[0m \u001b[38;5;28;01mexcept\u001b[39;00m \u001b[38;5;167;01mException\u001b[39;00m \u001b[38;5;28;01mas\u001b[39;00m e:\n\u001b[0;32m     67\u001b[0m     filtered_tb \u001b[38;5;241m=\u001b[39m _process_traceback_frames(e\u001b[38;5;241m.\u001b[39m__traceback__)\n",
      "File \u001b[1;32mf:\\KTU\\Neuroninių tinklų metodai\\venv\\Lib\\site-packages\\keras\\src\\engine\\training.py:1807\u001b[0m, in \u001b[0;36mModel.fit\u001b[1;34m(self, x, y, batch_size, epochs, verbose, callbacks, validation_split, validation_data, shuffle, class_weight, sample_weight, initial_epoch, steps_per_epoch, validation_steps, validation_batch_size, validation_freq, max_queue_size, workers, use_multiprocessing)\u001b[0m\n\u001b[0;32m   1799\u001b[0m \u001b[38;5;28;01mwith\u001b[39;00m tf\u001b[38;5;241m.\u001b[39mprofiler\u001b[38;5;241m.\u001b[39mexperimental\u001b[38;5;241m.\u001b[39mTrace(\n\u001b[0;32m   1800\u001b[0m     \u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mtrain\u001b[39m\u001b[38;5;124m\"\u001b[39m,\n\u001b[0;32m   1801\u001b[0m     epoch_num\u001b[38;5;241m=\u001b[39mepoch,\n\u001b[1;32m   (...)\u001b[0m\n\u001b[0;32m   1804\u001b[0m     _r\u001b[38;5;241m=\u001b[39m\u001b[38;5;241m1\u001b[39m,\n\u001b[0;32m   1805\u001b[0m ):\n\u001b[0;32m   1806\u001b[0m     callbacks\u001b[38;5;241m.\u001b[39mon_train_batch_begin(step)\n\u001b[1;32m-> 1807\u001b[0m     tmp_logs \u001b[38;5;241m=\u001b[39m \u001b[38;5;28;43mself\u001b[39;49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mtrain_function\u001b[49m\u001b[43m(\u001b[49m\u001b[43miterator\u001b[49m\u001b[43m)\u001b[49m\n\u001b[0;32m   1808\u001b[0m     \u001b[38;5;28;01mif\u001b[39;00m data_handler\u001b[38;5;241m.\u001b[39mshould_sync:\n\u001b[0;32m   1809\u001b[0m         context\u001b[38;5;241m.\u001b[39masync_wait()\n",
      "File \u001b[1;32mf:\\KTU\\Neuroninių tinklų metodai\\venv\\Lib\\site-packages\\tensorflow\\python\\util\\traceback_utils.py:150\u001b[0m, in \u001b[0;36mfilter_traceback.<locals>.error_handler\u001b[1;34m(*args, **kwargs)\u001b[0m\n\u001b[0;32m    148\u001b[0m filtered_tb \u001b[38;5;241m=\u001b[39m \u001b[38;5;28;01mNone\u001b[39;00m\n\u001b[0;32m    149\u001b[0m \u001b[38;5;28;01mtry\u001b[39;00m:\n\u001b[1;32m--> 150\u001b[0m   \u001b[38;5;28;01mreturn\u001b[39;00m \u001b[43mfn\u001b[49m\u001b[43m(\u001b[49m\u001b[38;5;241;43m*\u001b[39;49m\u001b[43margs\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[38;5;241;43m*\u001b[39;49m\u001b[38;5;241;43m*\u001b[39;49m\u001b[43mkwargs\u001b[49m\u001b[43m)\u001b[49m\n\u001b[0;32m    151\u001b[0m \u001b[38;5;28;01mexcept\u001b[39;00m \u001b[38;5;167;01mException\u001b[39;00m \u001b[38;5;28;01mas\u001b[39;00m e:\n\u001b[0;32m    152\u001b[0m   filtered_tb \u001b[38;5;241m=\u001b[39m _process_traceback_frames(e\u001b[38;5;241m.\u001b[39m__traceback__)\n",
      "File \u001b[1;32mf:\\KTU\\Neuroninių tinklų metodai\\venv\\Lib\\site-packages\\tensorflow\\python\\eager\\polymorphic_function\\polymorphic_function.py:832\u001b[0m, in \u001b[0;36mFunction.__call__\u001b[1;34m(self, *args, **kwds)\u001b[0m\n\u001b[0;32m    829\u001b[0m compiler \u001b[38;5;241m=\u001b[39m \u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mxla\u001b[39m\u001b[38;5;124m\"\u001b[39m \u001b[38;5;28;01mif\u001b[39;00m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39m_jit_compile \u001b[38;5;28;01melse\u001b[39;00m \u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mnonXla\u001b[39m\u001b[38;5;124m\"\u001b[39m\n\u001b[0;32m    831\u001b[0m \u001b[38;5;28;01mwith\u001b[39;00m OptionalXlaContext(\u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39m_jit_compile):\n\u001b[1;32m--> 832\u001b[0m   result \u001b[38;5;241m=\u001b[39m \u001b[38;5;28;43mself\u001b[39;49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43m_call\u001b[49m\u001b[43m(\u001b[49m\u001b[38;5;241;43m*\u001b[39;49m\u001b[43margs\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[38;5;241;43m*\u001b[39;49m\u001b[38;5;241;43m*\u001b[39;49m\u001b[43mkwds\u001b[49m\u001b[43m)\u001b[49m\n\u001b[0;32m    834\u001b[0m new_tracing_count \u001b[38;5;241m=\u001b[39m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39mexperimental_get_tracing_count()\n\u001b[0;32m    835\u001b[0m without_tracing \u001b[38;5;241m=\u001b[39m (tracing_count \u001b[38;5;241m==\u001b[39m new_tracing_count)\n",
      "File \u001b[1;32mf:\\KTU\\Neuroninių tinklų metodai\\venv\\Lib\\site-packages\\tensorflow\\python\\eager\\polymorphic_function\\polymorphic_function.py:868\u001b[0m, in \u001b[0;36mFunction._call\u001b[1;34m(self, *args, **kwds)\u001b[0m\n\u001b[0;32m    865\u001b[0m   \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39m_lock\u001b[38;5;241m.\u001b[39mrelease()\n\u001b[0;32m    866\u001b[0m   \u001b[38;5;66;03m# In this case we have created variables on the first call, so we run the\u001b[39;00m\n\u001b[0;32m    867\u001b[0m   \u001b[38;5;66;03m# defunned version which is guaranteed to never create variables.\u001b[39;00m\n\u001b[1;32m--> 868\u001b[0m   \u001b[38;5;28;01mreturn\u001b[39;00m \u001b[43mtracing_compilation\u001b[49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mcall_function\u001b[49m\u001b[43m(\u001b[49m\n\u001b[0;32m    869\u001b[0m \u001b[43m      \u001b[49m\u001b[43margs\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43mkwds\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[38;5;28;43mself\u001b[39;49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43m_no_variable_creation_config\u001b[49m\n\u001b[0;32m    870\u001b[0m \u001b[43m  \u001b[49m\u001b[43m)\u001b[49m\n\u001b[0;32m    871\u001b[0m \u001b[38;5;28;01melif\u001b[39;00m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39m_variable_creation_config \u001b[38;5;129;01mis\u001b[39;00m \u001b[38;5;129;01mnot\u001b[39;00m \u001b[38;5;28;01mNone\u001b[39;00m:\n\u001b[0;32m    872\u001b[0m   \u001b[38;5;66;03m# Release the lock early so that multiple threads can perform the call\u001b[39;00m\n\u001b[0;32m    873\u001b[0m   \u001b[38;5;66;03m# in parallel.\u001b[39;00m\n\u001b[0;32m    874\u001b[0m   \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39m_lock\u001b[38;5;241m.\u001b[39mrelease()\n",
      "File \u001b[1;32mf:\\KTU\\Neuroninių tinklų metodai\\venv\\Lib\\site-packages\\tensorflow\\python\\eager\\polymorphic_function\\tracing_compilation.py:139\u001b[0m, in \u001b[0;36mcall_function\u001b[1;34m(args, kwargs, tracing_options)\u001b[0m\n\u001b[0;32m    137\u001b[0m bound_args \u001b[38;5;241m=\u001b[39m function\u001b[38;5;241m.\u001b[39mfunction_type\u001b[38;5;241m.\u001b[39mbind(\u001b[38;5;241m*\u001b[39margs, \u001b[38;5;241m*\u001b[39m\u001b[38;5;241m*\u001b[39mkwargs)\n\u001b[0;32m    138\u001b[0m flat_inputs \u001b[38;5;241m=\u001b[39m function\u001b[38;5;241m.\u001b[39mfunction_type\u001b[38;5;241m.\u001b[39munpack_inputs(bound_args)\n\u001b[1;32m--> 139\u001b[0m \u001b[38;5;28;01mreturn\u001b[39;00m \u001b[43mfunction\u001b[49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43m_call_flat\u001b[49m\u001b[43m(\u001b[49m\u001b[43m  \u001b[49m\u001b[38;5;66;43;03m# pylint: disable=protected-access\u001b[39;49;00m\n\u001b[0;32m    140\u001b[0m \u001b[43m    \u001b[49m\u001b[43mflat_inputs\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43mcaptured_inputs\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[43mfunction\u001b[49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mcaptured_inputs\u001b[49m\n\u001b[0;32m    141\u001b[0m \u001b[43m\u001b[49m\u001b[43m)\u001b[49m\n",
      "File \u001b[1;32mf:\\KTU\\Neuroninių tinklų metodai\\venv\\Lib\\site-packages\\tensorflow\\python\\eager\\polymorphic_function\\concrete_function.py:1323\u001b[0m, in \u001b[0;36mConcreteFunction._call_flat\u001b[1;34m(self, tensor_inputs, captured_inputs)\u001b[0m\n\u001b[0;32m   1319\u001b[0m possible_gradient_type \u001b[38;5;241m=\u001b[39m gradients_util\u001b[38;5;241m.\u001b[39mPossibleTapeGradientTypes(args)\n\u001b[0;32m   1320\u001b[0m \u001b[38;5;28;01mif\u001b[39;00m (possible_gradient_type \u001b[38;5;241m==\u001b[39m gradients_util\u001b[38;5;241m.\u001b[39mPOSSIBLE_GRADIENT_TYPES_NONE\n\u001b[0;32m   1321\u001b[0m     \u001b[38;5;129;01mand\u001b[39;00m executing_eagerly):\n\u001b[0;32m   1322\u001b[0m   \u001b[38;5;66;03m# No tape is watching; skip to running the function.\u001b[39;00m\n\u001b[1;32m-> 1323\u001b[0m   \u001b[38;5;28;01mreturn\u001b[39;00m \u001b[38;5;28;43mself\u001b[39;49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43m_inference_function\u001b[49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mcall_preflattened\u001b[49m\u001b[43m(\u001b[49m\u001b[43margs\u001b[49m\u001b[43m)\u001b[49m\n\u001b[0;32m   1324\u001b[0m forward_backward \u001b[38;5;241m=\u001b[39m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39m_select_forward_and_backward_functions(\n\u001b[0;32m   1325\u001b[0m     args,\n\u001b[0;32m   1326\u001b[0m     possible_gradient_type,\n\u001b[0;32m   1327\u001b[0m     executing_eagerly)\n\u001b[0;32m   1328\u001b[0m forward_function, args_with_tangents \u001b[38;5;241m=\u001b[39m forward_backward\u001b[38;5;241m.\u001b[39mforward()\n",
      "File \u001b[1;32mf:\\KTU\\Neuroninių tinklų metodai\\venv\\Lib\\site-packages\\tensorflow\\python\\eager\\polymorphic_function\\atomic_function.py:216\u001b[0m, in \u001b[0;36mAtomicFunction.call_preflattened\u001b[1;34m(self, args)\u001b[0m\n\u001b[0;32m    214\u001b[0m \u001b[38;5;28;01mdef\u001b[39;00m \u001b[38;5;21mcall_preflattened\u001b[39m(\u001b[38;5;28mself\u001b[39m, args: Sequence[core\u001b[38;5;241m.\u001b[39mTensor]) \u001b[38;5;241m-\u001b[39m\u001b[38;5;241m>\u001b[39m Any:\n\u001b[0;32m    215\u001b[0m \u001b[38;5;250m  \u001b[39m\u001b[38;5;124;03m\"\"\"Calls with flattened tensor inputs and returns the structured output.\"\"\"\u001b[39;00m\n\u001b[1;32m--> 216\u001b[0m   flat_outputs \u001b[38;5;241m=\u001b[39m \u001b[38;5;28;43mself\u001b[39;49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mcall_flat\u001b[49m\u001b[43m(\u001b[49m\u001b[38;5;241;43m*\u001b[39;49m\u001b[43margs\u001b[49m\u001b[43m)\u001b[49m\n\u001b[0;32m    217\u001b[0m   \u001b[38;5;28;01mreturn\u001b[39;00m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39mfunction_type\u001b[38;5;241m.\u001b[39mpack_output(flat_outputs)\n",
      "File \u001b[1;32mf:\\KTU\\Neuroninių tinklų metodai\\venv\\Lib\\site-packages\\tensorflow\\python\\eager\\polymorphic_function\\atomic_function.py:251\u001b[0m, in \u001b[0;36mAtomicFunction.call_flat\u001b[1;34m(self, *args)\u001b[0m\n\u001b[0;32m    249\u001b[0m \u001b[38;5;28;01mwith\u001b[39;00m record\u001b[38;5;241m.\u001b[39mstop_recording():\n\u001b[0;32m    250\u001b[0m   \u001b[38;5;28;01mif\u001b[39;00m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39m_bound_context\u001b[38;5;241m.\u001b[39mexecuting_eagerly():\n\u001b[1;32m--> 251\u001b[0m     outputs \u001b[38;5;241m=\u001b[39m \u001b[38;5;28;43mself\u001b[39;49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43m_bound_context\u001b[49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mcall_function\u001b[49m\u001b[43m(\u001b[49m\n\u001b[0;32m    252\u001b[0m \u001b[43m        \u001b[49m\u001b[38;5;28;43mself\u001b[39;49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mname\u001b[49m\u001b[43m,\u001b[49m\n\u001b[0;32m    253\u001b[0m \u001b[43m        \u001b[49m\u001b[38;5;28;43mlist\u001b[39;49m\u001b[43m(\u001b[49m\u001b[43margs\u001b[49m\u001b[43m)\u001b[49m\u001b[43m,\u001b[49m\n\u001b[0;32m    254\u001b[0m \u001b[43m        \u001b[49m\u001b[38;5;28;43mlen\u001b[39;49m\u001b[43m(\u001b[49m\u001b[38;5;28;43mself\u001b[39;49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mfunction_type\u001b[49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mflat_outputs\u001b[49m\u001b[43m)\u001b[49m\u001b[43m,\u001b[49m\n\u001b[0;32m    255\u001b[0m \u001b[43m    \u001b[49m\u001b[43m)\u001b[49m\n\u001b[0;32m    256\u001b[0m   \u001b[38;5;28;01melse\u001b[39;00m:\n\u001b[0;32m    257\u001b[0m     outputs \u001b[38;5;241m=\u001b[39m make_call_op_in_graph(\n\u001b[0;32m    258\u001b[0m         \u001b[38;5;28mself\u001b[39m,\n\u001b[0;32m    259\u001b[0m         \u001b[38;5;28mlist\u001b[39m(args),\n\u001b[0;32m    260\u001b[0m         \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39m_bound_context\u001b[38;5;241m.\u001b[39mfunction_call_options\u001b[38;5;241m.\u001b[39mas_attrs(),\n\u001b[0;32m    261\u001b[0m     )\n",
      "File \u001b[1;32mf:\\KTU\\Neuroninių tinklų metodai\\venv\\Lib\\site-packages\\tensorflow\\python\\eager\\context.py:1486\u001b[0m, in \u001b[0;36mContext.call_function\u001b[1;34m(self, name, tensor_inputs, num_outputs)\u001b[0m\n\u001b[0;32m   1484\u001b[0m cancellation_context \u001b[38;5;241m=\u001b[39m cancellation\u001b[38;5;241m.\u001b[39mcontext()\n\u001b[0;32m   1485\u001b[0m \u001b[38;5;28;01mif\u001b[39;00m cancellation_context \u001b[38;5;129;01mis\u001b[39;00m \u001b[38;5;28;01mNone\u001b[39;00m:\n\u001b[1;32m-> 1486\u001b[0m   outputs \u001b[38;5;241m=\u001b[39m \u001b[43mexecute\u001b[49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mexecute\u001b[49m\u001b[43m(\u001b[49m\n\u001b[0;32m   1487\u001b[0m \u001b[43m      \u001b[49m\u001b[43mname\u001b[49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mdecode\u001b[49m\u001b[43m(\u001b[49m\u001b[38;5;124;43m\"\u001b[39;49m\u001b[38;5;124;43mutf-8\u001b[39;49m\u001b[38;5;124;43m\"\u001b[39;49m\u001b[43m)\u001b[49m\u001b[43m,\u001b[49m\n\u001b[0;32m   1488\u001b[0m \u001b[43m      \u001b[49m\u001b[43mnum_outputs\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[43mnum_outputs\u001b[49m\u001b[43m,\u001b[49m\n\u001b[0;32m   1489\u001b[0m \u001b[43m      \u001b[49m\u001b[43minputs\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[43mtensor_inputs\u001b[49m\u001b[43m,\u001b[49m\n\u001b[0;32m   1490\u001b[0m \u001b[43m      \u001b[49m\u001b[43mattrs\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[43mattrs\u001b[49m\u001b[43m,\u001b[49m\n\u001b[0;32m   1491\u001b[0m \u001b[43m      \u001b[49m\u001b[43mctx\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[38;5;28;43mself\u001b[39;49m\u001b[43m,\u001b[49m\n\u001b[0;32m   1492\u001b[0m \u001b[43m  \u001b[49m\u001b[43m)\u001b[49m\n\u001b[0;32m   1493\u001b[0m \u001b[38;5;28;01melse\u001b[39;00m:\n\u001b[0;32m   1494\u001b[0m   outputs \u001b[38;5;241m=\u001b[39m execute\u001b[38;5;241m.\u001b[39mexecute_with_cancellation(\n\u001b[0;32m   1495\u001b[0m       name\u001b[38;5;241m.\u001b[39mdecode(\u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mutf-8\u001b[39m\u001b[38;5;124m\"\u001b[39m),\n\u001b[0;32m   1496\u001b[0m       num_outputs\u001b[38;5;241m=\u001b[39mnum_outputs,\n\u001b[1;32m   (...)\u001b[0m\n\u001b[0;32m   1500\u001b[0m       cancellation_manager\u001b[38;5;241m=\u001b[39mcancellation_context,\n\u001b[0;32m   1501\u001b[0m   )\n",
      "File \u001b[1;32mf:\\KTU\\Neuroninių tinklų metodai\\venv\\Lib\\site-packages\\tensorflow\\python\\eager\\execute.py:53\u001b[0m, in \u001b[0;36mquick_execute\u001b[1;34m(op_name, num_outputs, inputs, attrs, ctx, name)\u001b[0m\n\u001b[0;32m     51\u001b[0m \u001b[38;5;28;01mtry\u001b[39;00m:\n\u001b[0;32m     52\u001b[0m   ctx\u001b[38;5;241m.\u001b[39mensure_initialized()\n\u001b[1;32m---> 53\u001b[0m   tensors \u001b[38;5;241m=\u001b[39m \u001b[43mpywrap_tfe\u001b[49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mTFE_Py_Execute\u001b[49m\u001b[43m(\u001b[49m\u001b[43mctx\u001b[49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43m_handle\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43mdevice_name\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43mop_name\u001b[49m\u001b[43m,\u001b[49m\n\u001b[0;32m     54\u001b[0m \u001b[43m                                      \u001b[49m\u001b[43minputs\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43mattrs\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43mnum_outputs\u001b[49m\u001b[43m)\u001b[49m\n\u001b[0;32m     55\u001b[0m \u001b[38;5;28;01mexcept\u001b[39;00m core\u001b[38;5;241m.\u001b[39m_NotOkStatusException \u001b[38;5;28;01mas\u001b[39;00m e:\n\u001b[0;32m     56\u001b[0m   \u001b[38;5;28;01mif\u001b[39;00m name \u001b[38;5;129;01mis\u001b[39;00m \u001b[38;5;129;01mnot\u001b[39;00m \u001b[38;5;28;01mNone\u001b[39;00m:\n",
      "\u001b[1;31mKeyboardInterrupt\u001b[0m: "
     ]
    }
   ],
   "source": [
    "#https://www.tensorflow.org/tutorials/images/segmentation\n",
    "\n",
    "import pix2pix\n",
    "\n",
    "(train_images0, train_labels0),(test_images0,test_labels0) = datasets.mnist.load_data()\n",
    "train_images0 = tf.pad(train_images0, [[0, 0], [2,2], [2,2]])/255\n",
    "test_images0 = tf.pad(test_images0, [[0, 0], [2,2], [2,2]])/255\n",
    "train_images0 = tf.expand_dims(train_images0, axis=3, name=None)\n",
    "test_images0 = tf.expand_dims(test_images0, axis=3, name=None)\n",
    "train_images0 = tf.repeat(train_images0, 3, axis=3)\n",
    "test_images0 = tf.repeat(test_images0, 3, axis=3)\n",
    "\n",
    "test_images  = test_images0\n",
    "train_images = train_images0\n",
    "\n",
    "base_model = tf.keras.applications.MobileNetV2(input_shape=[32, 32, 3], include_top=False)\n",
    "\n",
    "# Use the activations of these layers\n",
    "layer_names = [\n",
    "    'block_1_expand_relu',   # 64x64\n",
    "    'block_3_expand_relu',   # 32x32\n",
    "    'block_6_expand_relu',   # 16x16\n",
    "    'block_13_expand_relu',  # 8x8\n",
    "    'block_16_project',      # 4x4\n",
    "]\n",
    "base_model_outputs = [base_model.get_layer(name).output for name in layer_names]\n",
    "\n",
    "# Create the feature extraction model\n",
    "down_stack = tf.keras.Model(inputs=base_model.input, outputs=base_model_outputs)\n",
    "\n",
    "down_stack.trainable = False\n",
    "\n",
    "up_stack = [\n",
    "    pix2pix.upsample(512, 3),  # 4x4 -> 8x8\n",
    "    pix2pix.upsample(256, 3),  # 8x8 -> 16x16\n",
    "    pix2pix.upsample(128, 3),  # 16x16 -> 32x32\n",
    "    pix2pix.upsample(64, 3),   # 32x32 -> 64x64\n",
    "]\n",
    "\n",
    "def unet_model(output_channels:int):\n",
    "  inputs = tf.keras.layers.Input(shape=[32, 32, 3])\n",
    "\n",
    "  # Downsampling through the model\n",
    "  skips = down_stack(inputs)\n",
    "  x = skips[-1]\n",
    "  skips = reversed(skips[:-1])\n",
    "\n",
    "  # Upsampling and establishing the skip connections\n",
    "  for up, skip in zip(up_stack, skips):\n",
    "    x = up(x)\n",
    "    concat = tf.keras.layers.Concatenate()\n",
    "    x = concat([x, skip])\n",
    "\n",
    "  x = layers.Flatten()(x)\n",
    "  x = layers.Dense(128, activation='relu')(x)\n",
    "  x = layers.Dense(10, activation = 'softmax')(x)\n",
    "\n",
    "  return tf.keras.Model(inputs=inputs, outputs=x)\n",
    "\n",
    "model = unet_model(output_channels=10)\n",
    "model.compile(\n",
    "  optimizer='adam',\n",
    "  loss=tf.keras.losses.SparseCategoricalCrossentropy(from_logits=False),\n",
    "  metrics=['accuracy']\n",
    ")\n",
    "\n",
    "#tf.keras.utils.plot_model(model, show_shapes=True)\n",
    "\n",
    "model_history = model.fit(\n",
    "  train_images, train_labels0,\n",
    "  epochs=20,\n",
    "  batch_size=64,\n",
    "  validation_data=(test_images, test_labels0)\n",
    ")"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "venv",
   "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": "3.11.0"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}