model1-Conv_3-batchnormalisation-augmentation-dropout-1.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Using TensorFlow backend.\n"
     ]
    }
   ],
   "source": [
    "import tensorflow as tf\n",
    "import keras\n",
    "from keras.callbacks import *\n",
    "import numpy as np\n",
    "import pandas as pd\n",
    "import matplotlib.pyplot as plt\n",
    "from keras.models import Sequential\n",
    "from keras.layers import *"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "traindata = pd.read_csv('sign_mnist_train.csv')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "trainlabel=traindata['label'].values"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "traindata.drop('label',inplace=True,axis=1)\n",
    "trainimages = traindata.values"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "trainimages=trainimages.reshape(-1,28,28,1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "testdata = pd.read_csv('sign_mnist_test.csv')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "testlabel=testdata['label'].values"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "testdata.drop('label',inplace=True,axis=1)\n",
    "testimages = testdata.values"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "testimages=testimages.reshape(-1,28,28,1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "from keras.preprocessing.image import ImageDataGenerator"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [],
   "source": [
    "traingen=ImageDataGenerator(rotation_range=40,zoom_range=0.2,width_shift_range=0.2,height_shift_range=0.2,\n",
    "                  shear_range=0.2,horizontal_flip=True,rescale=1/255.0,validation_split=0.2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [],
   "source": [
    "traindata_generator = traingen.flow(trainimages,trainlabel,subset='training')\n",
    "validationdata_generator = traingen.flow(trainimages,trainlabel,subset='validation')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [],
   "source": [
    "testgen=ImageDataGenerator(rescale=1/255.0)\n",
    "testdata_generator = testgen.flow(testimages,testlabel)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [],
   "source": [
    "model=Sequential([])\n",
    "\n",
    "model.add(Conv2D(64,(3,3),activation=\"relu\",input_shape=(28,28,1)))\n",
    "model.add(BatchNormalization())\n",
    "# model.add(Conv2D(64,(3,3),activation=\"relu\"))\n",
    "# model.add(BatchNormalization())\n",
    "model.add(MaxPooling2D(2,2))\n",
    "#model.add(Dropout(0.4))\n",
    "\n",
    "model.add(Conv2D(128,(3,3),activation=\"relu\"))\n",
    "model.add(BatchNormalization())\n",
    "# model.add(Conv2D(128,(3,3),activation=\"relu\"))\n",
    "# model.add(BatchNormalization())\n",
    "model.add(MaxPooling2D(2,2))\n",
    "\n",
    "model.add(Conv2D(128,(3,3),activation=\"relu\"))\n",
    "model.add(BatchNormalization())\n",
    "# model.add(Conv2D(128,(3,3),activation=\"relu\"))\n",
    "# model.add(BatchNormalization())\n",
    "model.add(MaxPooling2D(2,2))\n",
    "\n",
    "model.add(Flatten())\n",
    "model.add(BatchNormalization())\n",
    "model.add(Dropout(0.4))\n",
    "model.add(Dense(256,activation=\"relu\"))\n",
    "model.add(BatchNormalization())\n",
    "model.add(Dense(26,activation=\"softmax\"))\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [],
   "source": [
    "model.compile(loss=\"sparse_categorical_crossentropy\",optimizer='adam',metrics=['accuracy'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model: \"sequential_1\"\n",
      "_________________________________________________________________\n",
      "Layer (type)                 Output Shape              Param #   \n",
      "=================================================================\n",
      "conv2d_1 (Conv2D)            (None, 26, 26, 64)        640       \n",
      "_________________________________________________________________\n",
      "batch_normalization_1 (Batch (None, 26, 26, 64)        256       \n",
      "_________________________________________________________________\n",
      "max_pooling2d_1 (MaxPooling2 (None, 13, 13, 64)        0         \n",
      "_________________________________________________________________\n",
      "conv2d_2 (Conv2D)            (None, 11, 11, 128)       73856     \n",
      "_________________________________________________________________\n",
      "batch_normalization_2 (Batch (None, 11, 11, 128)       512       \n",
      "_________________________________________________________________\n",
      "max_pooling2d_2 (MaxPooling2 (None, 5, 5, 128)         0         \n",
      "_________________________________________________________________\n",
      "conv2d_3 (Conv2D)            (None, 3, 3, 128)         147584    \n",
      "_________________________________________________________________\n",
      "batch_normalization_3 (Batch (None, 3, 3, 128)         512       \n",
      "_________________________________________________________________\n",
      "max_pooling2d_3 (MaxPooling2 (None, 1, 1, 128)         0         \n",
      "_________________________________________________________________\n",
      "flatten_1 (Flatten)          (None, 128)               0         \n",
      "_________________________________________________________________\n",
      "batch_normalization_4 (Batch (None, 128)               512       \n",
      "_________________________________________________________________\n",
      "dropout_1 (Dropout)          (None, 128)               0         \n",
      "_________________________________________________________________\n",
      "dense_1 (Dense)              (None, 256)               33024     \n",
      "_________________________________________________________________\n",
      "batch_normalization_5 (Batch (None, 256)               1024      \n",
      "_________________________________________________________________\n",
      "dense_2 (Dense)              (None, 26)                6682      \n",
      "=================================================================\n",
      "Total params: 264,602\n",
      "Trainable params: 263,194\n",
      "Non-trainable params: 1,408\n",
      "_________________________________________________________________\n"
     ]
    }
   ],
   "source": [
    "model.summary()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Define a Callback class that stops training once accuracy reaches 97.0%\n",
    "class myCallback(Callback):\n",
    "  def on_epoch_end(self, epoch, logs={}):\n",
    "    if(logs.get('accuracy')>0.985):\n",
    "      print(\"\\nReached 98.5% accuracy so cancelling training!\")\n",
    "      self.model.stop_training = True\n",
    "callback=myCallback()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/60\n",
      "687/687 [==============================] - 11s 16ms/step - loss: 2.5432 - accuracy: 0.2390 - val_loss: 3.4136 - val_accuracy: 0.1308\n",
      "Epoch 2/60\n",
      "687/687 [==============================] - 8s 12ms/step - loss: 1.7279 - accuracy: 0.4403 - val_loss: 4.1014 - val_accuracy: 0.2038\n",
      "Epoch 3/60\n",
      "687/687 [==============================] - 8s 12ms/step - loss: 1.4010 - accuracy: 0.5399 - val_loss: 1.4010 - val_accuracy: 0.6270\n",
      "Epoch 4/60\n",
      "687/687 [==============================] - 8s 12ms/step - loss: 1.1827 - accuracy: 0.6066 - val_loss: 1.6946 - val_accuracy: 0.4201\n",
      "Epoch 5/60\n",
      "687/687 [==============================] - 8s 12ms/step - loss: 1.0542 - accuracy: 0.6455 - val_loss: 2.4903 - val_accuracy: 0.4342\n",
      "Epoch 6/60\n",
      "687/687 [==============================] - 8s 12ms/step - loss: 0.9585 - accuracy: 0.6794 - val_loss: 2.5667 - val_accuracy: 0.5119\n",
      "Epoch 7/60\n",
      "687/687 [==============================] - 8s 12ms/step - loss: 0.8870 - accuracy: 0.7041 - val_loss: 1.2615 - val_accuracy: 0.6671\n",
      "Epoch 8/60\n",
      "687/687 [==============================] - 8s 12ms/step - loss: 0.8258 - accuracy: 0.7193 - val_loss: 0.9471 - val_accuracy: 0.6248\n",
      "Epoch 9/60\n",
      "687/687 [==============================] - 8s 12ms/step - loss: 0.7792 - accuracy: 0.7375 - val_loss: 0.6548 - val_accuracy: 0.7556\n",
      "Epoch 10/60\n",
      "687/687 [==============================] - 8s 12ms/step - loss: 0.7610 - accuracy: 0.7427 - val_loss: 0.5530 - val_accuracy: 0.7472\n",
      "Epoch 11/60\n",
      "687/687 [==============================] - 8s 12ms/step - loss: 0.7080 - accuracy: 0.7624 - val_loss: 0.3568 - val_accuracy: 0.7449\n",
      "Epoch 12/60\n",
      "687/687 [==============================] - 8s 12ms/step - loss: 0.6856 - accuracy: 0.7687 - val_loss: 0.5107 - val_accuracy: 0.7760\n",
      "Epoch 13/60\n",
      "687/687 [==============================] - 8s 12ms/step - loss: 0.6487 - accuracy: 0.7816 - val_loss: 0.4931 - val_accuracy: 0.7745\n",
      "Epoch 14/60\n",
      "687/687 [==============================] - 8s 12ms/step - loss: 0.6248 - accuracy: 0.7893 - val_loss: 0.9851 - val_accuracy: 0.7529\n",
      "Epoch 15/60\n",
      "687/687 [==============================] - 8s 12ms/step - loss: 0.6211 - accuracy: 0.7929 - val_loss: 1.4962 - val_accuracy: 0.6168\n",
      "Epoch 16/60\n",
      "687/687 [==============================] - 8s 12ms/step - loss: 0.5932 - accuracy: 0.7988 - val_loss: 1.0021 - val_accuracy: 0.7310\n",
      "Epoch 17/60\n",
      "687/687 [==============================] - 8s 12ms/step - loss: 0.5678 - accuracy: 0.8096 - val_loss: 0.6844 - val_accuracy: 0.7969\n",
      "Epoch 18/60\n",
      "687/687 [==============================] - 8s 12ms/step - loss: 0.5611 - accuracy: 0.8102 - val_loss: 1.8350 - val_accuracy: 0.6409\n",
      "Epoch 19/60\n",
      "687/687 [==============================] - 8s 12ms/step - loss: 0.5486 - accuracy: 0.8177 - val_loss: 0.5102 - val_accuracy: 0.8161\n",
      "Epoch 20/60\n",
      "687/687 [==============================] - 8s 12ms/step - loss: 0.5321 - accuracy: 0.8210 - val_loss: 2.2282 - val_accuracy: 0.5853\n",
      "Epoch 21/60\n",
      "687/687 [==============================] - 8s 12ms/step - loss: 0.5317 - accuracy: 0.8234 - val_loss: 0.8461 - val_accuracy: 0.7250\n",
      "Epoch 22/60\n",
      "687/687 [==============================] - 8s 12ms/step - loss: 0.5028 - accuracy: 0.8341 - val_loss: 1.1392 - val_accuracy: 0.6882\n",
      "Epoch 23/60\n",
      "687/687 [==============================] - 8s 12ms/step - loss: 0.4922 - accuracy: 0.8334 - val_loss: 0.4983 - val_accuracy: 0.7480\n",
      "Epoch 24/60\n",
      "687/687 [==============================] - 8s 12ms/step - loss: 0.4933 - accuracy: 0.8328 - val_loss: 1.2001 - val_accuracy: 0.6245\n",
      "Epoch 25/60\n",
      "687/687 [==============================] - 8s 12ms/step - loss: 0.4790 - accuracy: 0.8389 - val_loss: 0.3748 - val_accuracy: 0.8620\n",
      "Epoch 26/60\n",
      "687/687 [==============================] - 8s 12ms/step - loss: 0.4694 - accuracy: 0.8425 - val_loss: 0.5790 - val_accuracy: 0.8008\n",
      "Epoch 27/60\n",
      "687/687 [==============================] - 8s 12ms/step - loss: 0.4674 - accuracy: 0.8429 - val_loss: 0.2346 - val_accuracy: 0.7997\n",
      "Epoch 28/60\n",
      "687/687 [==============================] - 8s 12ms/step - loss: 0.4607 - accuracy: 0.8417 - val_loss: 0.3303 - val_accuracy: 0.8880\n",
      "Epoch 29/60\n",
      "687/687 [==============================] - 8s 12ms/step - loss: 0.4384 - accuracy: 0.8525 - val_loss: 0.7198 - val_accuracy: 0.8554\n",
      "Epoch 30/60\n",
      "687/687 [==============================] - 8s 12ms/step - loss: 0.4382 - accuracy: 0.8555 - val_loss: 0.3805 - val_accuracy: 0.8459\n",
      "Epoch 31/60\n",
      "687/687 [==============================] - 8s 12ms/step - loss: 0.4154 - accuracy: 0.8603 - val_loss: 0.2796 - val_accuracy: 0.8219\n",
      "Epoch 32/60\n",
      "687/687 [==============================] - 8s 12ms/step - loss: 0.4172 - accuracy: 0.8582 - val_loss: 0.6102 - val_accuracy: 0.8827\n",
      "Epoch 33/60\n",
      "687/687 [==============================] - 8s 12ms/step - loss: 0.4089 - accuracy: 0.8623 - val_loss: 0.1163 - val_accuracy: 0.8802\n",
      "Epoch 34/60\n",
      "687/687 [==============================] - 8s 12ms/step - loss: 0.4089 - accuracy: 0.8641 - val_loss: 1.1094 - val_accuracy: 0.7696\n",
      "Epoch 35/60\n",
      "687/687 [==============================] - 8s 12ms/step - loss: 0.4088 - accuracy: 0.8647 - val_loss: 0.2488 - val_accuracy: 0.7928\n",
      "Epoch 36/60\n",
      "687/687 [==============================] - 8s 12ms/step - loss: 0.4020 - accuracy: 0.8641 - val_loss: 0.1934 - val_accuracy: 0.8995\n",
      "Epoch 37/60\n",
      "687/687 [==============================] - 8s 12ms/step - loss: 0.3894 - accuracy: 0.8699 - val_loss: 0.3632 - val_accuracy: 0.9117\n",
      "Epoch 38/60\n",
      "687/687 [==============================] - 8s 12ms/step - loss: 0.3883 - accuracy: 0.8715 - val_loss: 0.2121 - val_accuracy: 0.8769\n",
      "Epoch 39/60\n",
      "687/687 [==============================] - 8s 12ms/step - loss: 0.3792 - accuracy: 0.8719 - val_loss: 0.2347 - val_accuracy: 0.8964\n",
      "Epoch 40/60\n",
      "687/687 [==============================] - 8s 12ms/step - loss: 0.3917 - accuracy: 0.8708 - val_loss: 0.8656 - val_accuracy: 0.8015\n",
      "Epoch 41/60\n",
      "687/687 [==============================] - 9s 13ms/step - loss: 0.3792 - accuracy: 0.8711 - val_loss: 0.7637 - val_accuracy: 0.8379\n",
      "Epoch 42/60\n",
      "687/687 [==============================] - 8s 12ms/step - loss: 0.3638 - accuracy: 0.8786 - val_loss: 0.3700 - val_accuracy: 0.8977\n",
      "Epoch 43/60\n",
      "687/687 [==============================] - 9s 13ms/step - loss: 0.3598 - accuracy: 0.8782 - val_loss: 0.4010 - val_accuracy: 0.7993\n",
      "Epoch 44/60\n",
      "687/687 [==============================] - 8s 12ms/step - loss: 0.3598 - accuracy: 0.8794 - val_loss: 0.5293 - val_accuracy: 0.8020\n",
      "Epoch 45/60\n",
      "687/687 [==============================] - 8s 12ms/step - loss: 0.3597 - accuracy: 0.8800 - val_loss: 0.3606 - val_accuracy: 0.8042\n",
      "Epoch 46/60\n",
      "687/687 [==============================] - 8s 12ms/step - loss: 0.3564 - accuracy: 0.8815 - val_loss: 0.2795 - val_accuracy: 0.9226\n",
      "Epoch 47/60\n",
      "687/687 [==============================] - 8s 12ms/step - loss: 0.3430 - accuracy: 0.8860 - val_loss: 0.5296 - val_accuracy: 0.8379\n",
      "Epoch 48/60\n",
      "687/687 [==============================] - 8s 12ms/step - loss: 0.3368 - accuracy: 0.8868 - val_loss: 0.3394 - val_accuracy: 0.9122\n",
      "Epoch 49/60\n",
      "687/687 [==============================] - 8s 12ms/step - loss: 0.3432 - accuracy: 0.8853 - val_loss: 0.3486 - val_accuracy: 0.8845\n",
      "Epoch 50/60\n",
      "687/687 [==============================] - 8s 12ms/step - loss: 0.3356 - accuracy: 0.8885 - val_loss: 0.4220 - val_accuracy: 0.8430\n",
      "Epoch 51/60\n",
      "687/687 [==============================] - 8s 12ms/step - loss: 0.3424 - accuracy: 0.8843 - val_loss: 0.4412 - val_accuracy: 0.8907\n",
      "Epoch 52/60\n",
      "687/687 [==============================] - 8s 12ms/step - loss: 0.3307 - accuracy: 0.8862 - val_loss: 0.2266 - val_accuracy: 0.8742\n",
      "Epoch 53/60\n",
      "687/687 [==============================] - 8s 12ms/step - loss: 0.3315 - accuracy: 0.8872 - val_loss: 0.3253 - val_accuracy: 0.8887\n",
      "Epoch 54/60\n",
      "687/687 [==============================] - 8s 12ms/step - loss: 0.3285 - accuracy: 0.8900 - val_loss: 0.3597 - val_accuracy: 0.8829\n",
      "Epoch 55/60\n",
      "687/687 [==============================] - 8s 12ms/step - loss: 0.3288 - accuracy: 0.8895 - val_loss: 0.1150 - val_accuracy: 0.8621\n",
      "Epoch 56/60\n",
      "687/687 [==============================] - 8s 12ms/step - loss: 0.3209 - accuracy: 0.8925 - val_loss: 0.1809 - val_accuracy: 0.8148\n",
      "Epoch 57/60\n",
      "687/687 [==============================] - 8s 12ms/step - loss: 0.3296 - accuracy: 0.8917 - val_loss: 0.2081 - val_accuracy: 0.8824\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 58/60\n",
      "687/687 [==============================] - 8s 12ms/step - loss: 0.3147 - accuracy: 0.8948 - val_loss: 0.0887 - val_accuracy: 0.9048\n",
      "Epoch 59/60\n",
      "687/687 [==============================] - 8s 12ms/step - loss: 0.3271 - accuracy: 0.8883 - val_loss: 0.1151 - val_accuracy: 0.8911\n",
      "Epoch 60/60\n",
      "687/687 [==============================] - 8s 12ms/step - loss: 0.3100 - accuracy: 0.8958 - val_loss: 0.2291 - val_accuracy: 0.9080\n"
     ]
    }
   ],
   "source": [
    "history=model.fit(traindata_generator,epochs=60,validation_data=validationdata_generator,callbacks=[callback])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAXQAAAEICAYAAABPgw/pAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4xLjEsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy8QZhcZAAAgAElEQVR4nOydd3hUZfbHP4fQe+9dQaQGCKCCgvWHZbGL2Atiw7aWxbWurq6uurq6rqvrimUVxIKCi6AgFlSk9y4ChtClhJ6Q8/vjzCSTYSaZhAnJhPN5nnlm7r3vfe+5M8l3zpz3vOcVVcVxHMdJfMoUtwGO4zhOfHBBdxzHKSW4oDuO45QSXNAdx3FKCS7ojuM4pQQXdMdxnFKCC3opRkSSRGSniDSPZ9viRESOFpG459qKyGkisipke6mInBhL20Jc63UR+WNhz3ecaJQtbgOcHERkZ8hmZWAfcCCwfaOqvluQ/lT1AFA13m2PBFT1mHj0IyKDgStUtV9I34Pj0bfjhOOCXoJQ1WxBDXiAg1V1YrT2IlJWVTMPh22Okx/+91j8eMglgRCRP4vI+yIyQkTSgStE5HgRmSoi20RknYi8KCLlAu3LioiKSMvA9n8Dxz8XkXQR+VFEWhW0beD4mSKyTES2i8hLIvK9iFwTxe5YbLxRRFaIyFYReTHk3CQReV5EtojIz0D/PN6fB0VkZNi+l0Xkb4HXg0VkceB+fg54z9H6ShWRfoHXlUXknYBtC4HuEa67MtDvQhEZENjfCfgHcGIgnLU55L19NOT8mwL3vkVEPhGRRrG8NwV5n4P2iMhEEflNRNaLyH0h13ko8J7sEJEZItI4UnhLRKYEP+fA+/lt4Dq/AQ+KSBsRmRy4l82B961GyPktAve4KXD87yJSMWDzsSHtGonIbhGpE+1+nQioqj9K4ANYBZwWtu/PwH7gd9iXcSWgB9AL+7XVGlgGDA20Lwso0DKw/V9gM5AClAPeB/5biLb1gXTg3MCx3wMZwDVR7iUWGz8FagAtgd+C9w4MBRYCTYE6wLf2ZxvxOq2BnUCVkL43AimB7d8F2ghwCrAH6Bw4dhqwKqSvVKBf4PWzwNdALaAFsCis7SVAo8BnclnAhgaBY4OBr8Ps/C/waOD1GQEbk4GKwD+Br2J5bwr4PtcANgB3ABWA6kDPwLH7gblAm8A9JAO1gaPD32tgSvBzDtxbJnAzkIT9PbYFTgXKB/5OvgeeDbmfBYH3s0qgfe/AsdeAJ0Kuczcwurj/DxPtUewG+CPKBxNd0L/K57x7gA8CryOJ9L9C2g4AFhSi7XXAdyHHBFhHFEGP0cbjQo5/DNwTeP0tFnoKHjsrXGTC+p4KXBZ4fSawLI+2nwG3Bl7nJehrQj8L4JbQthH6XQCcHXidn6C/BTwZcqw6Nm7SNL/3poDv85XAjCjtfg7aG7Y/FkFfmY8NFwHTA69PBNYDSRHa9QZ+ASSwPQe4IN7/V6X94SGXxOPX0A0RaSci/wv8hN4BPAbUzeP89SGvd5P3QGi0to1D7VD7D0yN1kmMNsZ0LWB1HvYCvAcMCry+DMgeSBaRc0Tkp0DIYRvmHef1XgVplJcNInKNiMwNhA22Ae1i7Bfs/rL7U9UdwFagSUibmD6zfN7nZsCKKDY0w0S9MIT/PTYUkVEisjZgw5thNqxSG4DPhap+j3n7fUSkI9Ac+F8hbTpicUFPPMJT9l7FPMKjVbU68DDmMRcl6zAPEgAREXILUDiHYuM6TAiC5JdW+T5wmog0xUJC7wVsrAR8CPwFC4fUBL6I0Y710WwQkdbAK1jYoU6g3yUh/eaXYpmGhXGC/VXDQjtrY7ArnLze51+Bo6KcF+3YroBNlUP2NQxrE35/T2PZWZ0CNlwTZkMLEUmKYsfbwBXYr4lRqrovSjsnCi7oiU81YDuwKzCodONhuOZnQDcR+Z2IlMXisvWKyMZRwJ0i0iQwQPaHvBqr6gYsLDAcWKqqywOHKmBx3U3AARE5B4v1xmrDH0Wkplie/tCQY1UxUduEfbcNxjz0IBuApqGDk2GMAK4Xkc4iUgH7wvlOVaP+4smDvN7nMUBzERkqIuVFpLqI9Awcex34s4gcJUayiNTGvsjWY4PvSSIyhJAvnzxs2AVsF5FmWNgnyI/AFuBJsYHmSiLSO+T4O1iI5jJM3J0C4oKe+NwNXI0NUr6KeahFSkA0BwJ/w/5BjwJmY55ZvG18BZgEzAemY152fryHxcTfC7F5G3AXMBobWLwI+2KKhUewXwqrgM8JERtVnQe8CEwLtGkH/BRy7pfAcmCDiISGToLnj8dCI6MD5zcHLo/RrnCivs+quh04HbgQG4RdBvQNHH4G+AR7n3dgA5QVA6G0G4A/YgPkR4fdWyQeAXpiXyxjgI9CbMgEzgGOxbz1NdjnEDy+Cvuc96vqDwW8d4ecAQjHKTSBn9BpwEWq+l1x2+MkLiLyNjbQ+mhx25KI+MQip1CISH/sJ/ReLO0tE/NSHadQBMYjzgU6FbctiYqHXJzC0gdYif0U7w+c54NYTmERkb9gufBPquqa4rYnUfGQi+M4TinBPXTHcZxSQrHF0OvWrastW7Ysrss7juMkJDNnztysqhHThItN0Fu2bMmMGTOK6/KO4zgJiYhEnS3tIRfHcZxSggu64zhOKcEF3XEcp5Tggu44jlNKcEF3HMcpJbigO47jlBJc0B3HcUoJLuiO48SNPXvg9dchK6u4LTkycUF3HCdujBoFN9wAkyYVtyVHJi7ojuPEjXnz7Pnbb4vXjiMVF3THKWFs2gS33QZ//WtxW1Jw5s+352++ybvdgQPQrx+89lqRm1Ti2Lu36PqOSdBFpL+ILBWRFSIyLMLxFiIySUTmicjXgQV6HccpAJmZ8NJL0LYt/OMfMGwYJFq5owUL7Pmnn/IWrhkzTPSHDoXp0w+PbUVNLOMGc+ZAu3bw8cdFY0O+gh5YXuxl4EygPTBIRNqHNXsWeFtVO2PrI/4l3oY6Tmnmm2+gWze4/XZISYEff4T69U3wEmWAccsWWLcOTjoJ9u83UY/GhAkgAg0awCWXwLZth8/OouDxx6FmTfj736N/Xh98AL1726+Toio0G4uH3hNYoaorVXU/MBJbJiqU9tgCswCTIxx3HCcKd91l4YcdO+Cjj+CLL+C44yzk8tNP8NZbRXftX36BnTvj01cw3HLzzSbWecXRJ0ywL64PP4TUVLj+ekjUtXZeegkefhhq14Y777QvtGXLco5nZcEjj9gXV3Ky/Trp1rWIblZV83xgq3K/HrJ9JfCPsDbvAXcEXl8AKFAnQl9DgBnAjObNm6vjHOns2aOalKR68cWqu3blPnbggOoJJ6jWq6e6dWvk8595RrVXr4PPjYX0dNVq1VTvvrvg50bipZdUQTU1VbVLF9VTT43c7rffVMuUUX3wQdt+9lk776WX4mPHYWPfPv3v06kKqud1WKoZwx7Ut+5frDVrZmnFinZf27apnn++3d+1l+7SvU+/oJqcrDp2bKEvC8zQaHod7YDmiPDFEQT9pbA2jYGPgdnA34FUoEZe/Xbv3r3QN+Q4pYXp0+2/8MMPIx+fNUtVRPX223Pvz8pSvf9+OxdUv/664Nd+9107t1evgp8biSFDVGvVMttuv121UiXVffsObvfBB3bd776z7aws1XPOUS1fXnXGjNiutXy56iuvqGZk5N82KyvkxcaNqj/8YDf/9tuq77+v+sknquPGqU6cqPree6p/+pPqlVeqHnecat26qhUqqNasqdqokWrr1qodOqgedZT+T87WsuzXfnyle6hg31KgabU76LnNZymoVq6cpWXkgL7Q7p+aJXZcU1JUP/88thuNQF6CHssCF6lAs5DtpkBamJefFvDMEZGqwIWquj3WXwmOUxRs2QLnnQfPPQc9exa3NZGZNcueu3WLfLxrV7jpJhskvf566NzZfsLffju8/DJcfjm8+y5MnQp9+xbs2iNG2PPs2RbzLl++8PcBFnLp1MnCLSedBC++CLNmZHFczSV2ke3bYfduJow8heoV2tPrgwdhQiWkdm3e7N+IrlPP5ZIBMOu5r6mRuhCWLrXYxdKl9r3VrZvFabp359bn+/PFtxX5cEQG7z/1C3X2pcGGDZYilJ4OO3eSuWM3T31/Ik/OO5u25VfR/8D/6L9/DCfwA+XJyPtmmjaFNm3gggssOL5nT67HlG0duXD1g3RptoNPX69GxW7rISkJxo+n0ejRjP6sHyM5ixd3385jPMzpu5fB/cPgyittVLSoiKb0muN9l8VWd28FlMdW5u4Q1qYuUCbw+gngsfz6dQ/dKWpGjjSH6NhjVffujU+fixapPvSQ6quvqk6YoLpkieru3YXv76abzPnL9iIjsGWLap06qiedpLp/vzmPoHrvvXbe0Uernndewa67ZYtquXJ2LtgvhUMha+curVb1gN5y8UbVsWN1w++fUlB9quLDOT8jQLNAm7FaL5CPVKtWzfZqFfR7jtckMvQ6Xrd99eqp9umjev31qtdeq9q5s2pSki6inYLq6Xyh5dmrLVmpc+ic6zorpbX2TvpBQfWcqpO1X+05WrZMpoJqtUr79bxTd+i8z1arLligOnOmee2TJ6vOn59v/Gr+fNUaNVTbtjWHPyJ795rX//DD9vPpwIFDe4ND4FBCLnY+ZwHLgJ+BBwL7HgMGaE6cfXmgzetAhfz6dEF3ipqhQy0+DaqPPRafPs87L5duZD/atlXdvr3g/fXsqXryyfm3e/VVu06nTvb85z/nfAlccYVqw4Z5fymE8+9/Wz8fXf2pgurLZ32m+sQTqn/9q+rzz6s+/bTqAw/Ym3jllarnnmtB8T59LGTQqZPddJMmqhUr6iqaK6i+wo3WcZkyemyFn/Ws5vNUhw834Vy/XhdN36lg96OqJnRbt6r+/LPq9Ol65/mrNCkpS5dPjzJosGuX3nT+eq1QNkM33v64Tr3jPW1ca5dWqpCpI59ZrVnrN+ib/96n1aplafXqqv/9b86p27erjh6teuONFkmpXVt1zpzY37MgF11kX7CrVxf83HhwyIJeFA8XdKeoSU42Dbr0UovPLl58aP1t327h1KFDVdesUf32W9V33lG9+Wb7T5o0qWD9ZWSoVqyo+vvf5982M9N0FFRffDH3sX/8w/bHJDA7dqi++aaeWmuGHs0yzQKtz3q9hjcO/pZKSrKgeMuWNsp5wgmqp5yieuaZ9s02cKB5zvfco2OvGqWgOuWJr1W//141PV1vvFG1enWzPcjzz1vXq1ZFNm/dOntPrrkm8vHfflOtXNkuG3pO79454WmwXzPRrqGqumKFatOmJswFEfW9e+2HxY03xn5OvMlL0IttkWjHKUp27LBp6A89ZGl048fDkCHw9ddQppDzoz/9FPbtg8sug2bN7HHiiXDGGfDKKzZp5JRTYu9vyRKbfHNQ/DwjA6ZNswabN8PmzSRt3syYWlms7FKV3iNmw1v7rd3+/fTa0wH4kKlnPUbzNrOhenWoUsVk+cABe2Rm2pvyxRes312NyazlgZOmIG+soOedDZi28hqYdZn1mZFhAfXKlS0gHgMLnrLnDrf0hZr2um9fePVVmDs35x7Hj4djjoEWLSL307ChjRm89BI8+CAcdVTu46+/Drt3wx135D7nq68sZfA//4G//AXuvddC2tE46ij7W+jXD0491c7v3Dn/+5w82dI8BwzIv22xEE3pi/rhHrpTlEyYYJ7al1/a9uuBsOy//134Ps8+W7V588ihjcaNLTIRlawsS/lIT7dcxQMH9K23zKZFi1T1l19U//Uvy3GrXj23p1yhgrmTXbqY63n66apnnWVe8iWX6L4LB2nFMnv1941HWCikRQtzPevVs1hM06a2r00b1Ztu0hfv/FlBdeFCM+2xxyyTpjAhoyCXXabarFnufb/+aub/7W+2vXu3ed/hGTvhpKVZu1AvXNV+0TRvrtq3b/RzI2XV5MXy5RY1qltXdd68/NvffLNqlSr2ERYXeMjFSVTWrg0IXgF56CGLGOzYYdtZWar9+tkA5Lp1Be/vt99sEPGee8IOrFun+tlneuYxK7Rz3bWmbKeeqtqxo2qDBibOFSqYYoaFNO6UF7QyOzWzRu2c/c2bq95wg+UxrlqlunNnTMHx3r0tIhILxx9v44tBPv/cLv3VV7G/H+F06mTfMeG0bp0zYBv8kh03Lv/+7rjDPr+ff87Z99FHdv7HHxfezkgsX25fyHXrWqg/GllZJv4XXBDf6xeUvATdQy5OieYPf7CZk2vXQtkC/LVOmWKz8qpVs20R+/nfubP9XH///QgnqcLy5fD99zmPrVuhcWNGH7iajIw7GLj9NXg0zfINZ86ENMvgTeYJvuRe9v04iwoNa8HRR8Pxx1vYokKFnEf58hYC2bePWf8ZQJfMTSRdeJm1/7//s3hEjGGOUI47zlIb80s/XLXKygo8+WTOvh497HnaNDj55AJfmowMiw6deebBx/r2tVBVVpbNDq1QIbb0yvvug3/9C554wsIoYNPqW7SIf7jj6KMt/NKnD9xyS/TCYrNm2d9hiQ23gHvoTsmmb1/zyr74IvZz9u+3SS2Rfto//rj196e7ftNtH0+yEcU77rCBvnr1cjzlWrUsxjJ4sOrZZ+sZ1X7Q1km/aBZi3vaxx1p6yfPPq37zjb4/fJeCTQSKhQMHbJbmLbfEfl95MWqUmT1tWt7tnnrK2q1cmXv/UUepXnhh4a69YIH1+c47Bx8bPtyOzZ+v2r696mmnxd7v7berli1rXvrs2dbPM88UzsZYeOYZzfMzfPhhy7LctKnobIgFPOTiJCrHHmt/pYMHx37OTz/ZOaNGhR3YtUv3PfGMnlNuvOUjs13v5WlNrXiUxaevukr1tdcsuBySN7xxo/38v/9+tW+LCInnS5faNd94IzYbly+39q+/Hvt95cWaNRrT9PnkZJsAGc6gQQfHwGNlxAi7dqRskZUr7diwYQUX5LVrLVp1/fWW9VK5soW+iorffrP4+NVXRz6enKx64olFd/1YcUF3EpbagfBy7dqmpbHw3HN2ztq1gR1795rSNWxoB/r31xkPfKyXnrJey5TJ0nLlsvTqqw/2WoP861922uzZ0a+ZmWlikN+AX5D337c+Z86MrX1+ZGVZHPjyy6O3WbzYrvnCCwcfC6YTpqUV/Np//KN94UWavJWVZWOylStb/7EMPIYydKh56eXL24BkUXPrrXat9etz71+1quh/IcRKXoLuMXSnxJKRAb/9Zilvs2bBpLG76d9ors0xX73aAsbBNLuMDAvUVqvGlPE30LpOYxp//hHs2gV/+5u1P+kkWyPtxBPpDowA/rIKnn/e0uEmTbLqho0b57bj/fetRnmXLtFtTUqy+PycObHd2+zZUK4cdOhQuPcmHBGLo0+dGr3NyJHW7pJLDj4WjKNPn17wGPGCBfb+VKgQ2a6+fa08QePG0LFjwfoeNswWwdi/38odFDXBkgr/+pdVSAwydqw9n1vS68hGU/qifriH7uRJVpamTVyooPp8+9e0hmzPPfklKckC5dWrW4pew4aqjRtrVtVqWo8NehVv5rRNSbEUizyyRWbPNg+7a1fLLAyybp3FTR96KH+Tb77ZpoTHMmPzjDPsWvHk6aftdiPFeLOybGJntFmpu3bZW/rAAwW/bqtWqpdcEv14cJZreBpirDz1lOpttxXu3MJw9tmq9evn/sVx+umq7dodPhvyAvfQnWJj924r8r13r2V8VKqU81yxYk7mR/B53jz47DP43//Y+GstYC7NdyzgvBZtGb1+EK/+ty7lu3W0dIcIM4SWL4NNx0CfFy6G8/rZTKA2bfLNHElONk98wACbODR6tHndH35ojv/AgfnfapcuNsFo9eq8FzBQtV8c8c6WOO44e/7pJzj77NzHPv/c6lzdd1/kcytXtsJaBV09KD3daqpfd130Nqefbh/thRcWrO8gf/hD4c4rLHfcYZPFRo6Eq6+2mmJff21160s80ZS+qB/uoZdywuPWBXlUqaJ63nk64a7PNVhm9X//s0P5lZF+I+DEFyZ3XTVnGv0dd9h2nz5WLTUWpk61c0ePzrtdcMLNP/5ROBujsXOnednBOuNB9u8377JNm7wn3gwZkn+hsHBivefQXz0lnawsy8jp2tVeB4u8TZlS3JYZuIfuHDYyM+Htt+FPf4I1a2xu/IgRluy7e7eVH9292x779tlj//6c1y1aWNC1QgU2/te6rF/fyt/WqmUh8HPOiX75KVOgTp3CVyi99VZYsQJeeMF+REyZAo89Ftu5nTrZj4Y5c6xsbzSCJXO7di2cjdGoUsXi+OFx9FdftTzxMWPyzlHv0cPi1StW2I+aWAiuUtSpU97tqlaNrb+SgIiVERgyBL77zt63evVyfgGVZFzQncKxezf8/LOJduhj6lRYudLU4d//tt/bhZgoA7Bxoz3Xr29CdP75ti7j3r0WrYnElCm2bmMhLwnAs8/aLTwVqE8SS7gFLGzRtm3+A6OzZ5t9eQ2yFpZeveC99yxMVKaMzYt65BGrV5LXFyHk1IyfNq1ggl65MrRqdWh2lzSuuALuvx+eecb+ps4/P+/aMCUFF3Qnf5Yvt6pES5bA4sX2WL06d5ty5axa1VFHWVbJgAGHpqrYegXly0ONGrY9cCC88YbNOIyUbbBxo8WJBw8+pMuSlGSieOqpFtpv2zb2c5OTbSZmXsyaZRNCq1Q5NDsjcdxxlqGxZAm0b2+/LrZts48kv4+jfXsT5+nTbeGMWJg/3zJ1ClvwrKRSqRLceGPOjNoSPTs0BBd0JzKbN9uo0Dvv8Ma0DnzLSbxZ6V8Wy+jd25bPCZbNa97clm+P83/1xo3mnQeF6OSTLZzy/vuRBf377+25T59Dv3aVKvDDD5YNWRCSk+1t27rVQkSRmD3bIlFFQTAsMHWqlUoIXekoP8qWhe7dzUOPlQUL4He/K5ytJZ1bbrGFusuWtR+aiYAL+pHM/PkWNtm712Lbe/daKOWbbywtIjMTOndmbIf7Gb/iKIanX4UkHT5XLCjoQcqVsxXB3nvPzK1UKXf7KVMsFBNtObeCUqZM5NzqvEhOtue5c600azibNsGvv8bPxnDatLEvkp9+gk8+sffo8cdjP79nT/sSyMiw9zsvgiu+5Rc/T1SaNIG777b3oih+TRUFMQm6iPTHFn9OwhaMfirseHPgLawSchIwTFXHxdlWJ17Mn2+Fwj/9NPLxRo1sVOjKK6FzZ9Z0h737IH2Xldo+XGzcaI5/KAMHWmj+889N3EOZMsUEqaAiHE+CcfFogj57tj3He0A0SJky9h68/76l2/3lLwe/h3nRo4eNTc+fn/+XTqwDoonMU0/l36Ykka+gi0gS8DJwOrZg9HQRGaOqi0KaPQiMUtVXRKQ9MA5oWQT2OofCihU2QjZihCnzn/8MZ52VkxMefK5WLVf4JBgu37jx8Ar6hg0Hz6Ts29cyDt57z3Ktg+K9a5fFpu+99/DZF4mGDU1Aow2MFrWgg4VdJkywXPg77yzYuaEDo1272kzdlSvtsXq1VRtcu9aKTP78s7Ut6OxPp+iIxUPvCaxQ1ZUAIjISOBcIFXQFgv/qNYC0eBrpHAK7d9uA5kcfWTphhQo2n/qee9hTqTbly+c9er9rF2zZYq83bLDsw8OB6sEhF7B45kUX2QSejz6ysEC1avY9lJkZn/j5oZKcHF3QZ80yoY0WX48H/fpZ1ugzz0TPBopGy5ZQty488IBN6NmxI/fxqlUtFNG0KfTvbx59QX4BOEVLLILeBPg1ZDsV6BXW5lHgCxG5DagCnBapIxEZAgwBaN68eUFtdWIhK8tSHCZMsLjEt9/ab+jKlW0ttgcegIYNOXAAOra19Kw//Sl6d7+GfPIbNhS9+UHS083scEEHs7djRwsp7NhhbXfssIyYwtTzjjfJyZZVEqk2+ezZRRc/D9Kvn83ezGu2ajRETMgnT4bWrXM/WrQ4vL/QnIITi6BHSnbSsO1BwJuq+pyIHA+8IyIdVTUr10mqrwGvAaSkpIT34RSUPXts5sP8+ZZusGABLFpkXjlYRsott9jKAyeemMtd+/FH+xk9c2bel1izJuf14RT04LUieX/16tltlVSSk20gbfHi3LnmU6daBujVVxe9DYUR8yD33GMPJ/GIRdBTgWYh2005OKRyPdAfQFV/FJGKQF1gYzyMdEJQtXy6t96yka/gb+KGDc1tHTLERqlOOSXP/+pPPrHnlSvzvlxxCXropKJEI5jpMmdOjqAvXWoTe446yj4ixykKYhH06UAbEWkFrAUuBS4La7MGOBV4U0SOBSoCm+Jp6BHPxo2W3vHWW+bmVa5sweTLLrPk4bp1Y+5KNUfQf/klZ1ZhJFavtmPVq7ugx0qbNja+PGeOeePr1lm8uUwZW/W+Xr3ittApreQr6KqaKSJDgQlYSuIbqrpQRB7DisSMAe4G/i0id2HhmGsCRWScQyUryxZVvO8+m/LXrx/88Y9Wui64YGYBWbjQMhSCg3fr1x9cAzzImjU2CFatmgt6rCQl2Y+kOXPsB9RZZ9n9fP314RtUdo5MYspDD+SUjwvb93DI60VA7/iaduQybJhlFtauvIfa6xZRe3tNajd6i5Q7Uxj8cONDnVHP6NE5BYiuucbCLnkJeosWlk1SHDH0RPVmk5OtkNiFF9oQx9ixOYtIOE5RUcoqMCQ+Y8bA009Da1lJs6WT2Lcrg4WNTufjjN8x5NHGPPTQoV/jk08sV/mEE2w7rzj6mjU5M/sPt4deq1be1QFLMsnJ9oNq4kRbDenMM4vbIudIwAW9pLBtG7/9+yNuHLiVzmXmM2F1O8Zc+QFT0o5iUVpNNm4UBg+GJ56w2X+FZc0ay4U+7zzzvEUsjh6JrCxLWywuQU/EcEuQ44+35yeesF9BjnM48Fouxc2IEebCffstd2W+ziaqMe680ZT//aRcFZxErIre7t0WQq9aFW67reCXCw6Gnn++eb9Nm0b30Nevt/S75s3N29y5065fuXIh7rOAbNiQ2BNWkpPt/Uvke3ASD/fQiwtVm3p/2WWQmspn573O21zNHx9MouvohyOW40tKgjffNO/69tutlGxB+eQTK5MarHfdunV0QdjtOH8AACAASURBVA+mLAY9dDh8Xnqie+jgYu4cflzQi4Bt2yyz4U9/gtTUCA1UbeTzoYfgqqvYOmUhN/5wNR07woMP5T3iWa6clWc94wyr+z1yZOx2bdliE0dDV9OJRdBbtHBBd5xEwAW9CPjhB5t1/+ijJobnnWfbBw5ggemhQ63Q8s03w/Dh/P6+smzYAMOHxzYIWKGCZar06QNXXWWly2Phf/8zG84/P2df69ZWaGnPnoPbh3roQXHdeBimimVkWFEoF3THKRgu6EXA4sX2PG2apY//8IN57EcfrVx+9E8M+2czXj71Y8b2f5k33izDm29a/YyUlNivUbkyPPigid+CBbGdM3q05ZR3756zr3Vre1616uD2a9bYakHVq8fXQ8/KskqJo0ZFPr4pMCXNQxaOUzB8ULQIWLLE8qd79LDHn+74jU+eXcGbbws/bmrAB0n3kjEpCSZZ+/bt4eGH8+4zEsccY89Ll0auvR3K7t1Wr+u663IvRRYU9JUr4dhjc5+zerV555DjLcdD0KdMgXHj7IvikksOPp7Ik4ocpzhxQS8CFi9W2jXYCnc9DpMnU37ePC5R5ZLKleGFJ8m67Q42bLCUwNRUE/3CLMrQrJlNMV+2LP+2X35pYZXQcAvkFvRwgjnoYHW9atSIj6AH4/4LF0Y+7oLuOIXDBT3eZGSwZMZeLtz3Aaz4l83eeewxq+vaoweUL08ZbFGgRo1yFhQoDGXKWLbK0qX5t/3kE6hZE046Kff+evVsea1ogh6cfATxyUXPyIAPPrDXS5daDfOyYX+FeVVadBwnOi7o8WTnTjadO5gt+0bS7rRmMHZrwVcYKCDHHJOzCk40MjNt6vk55xy8TqQItGp1sKDv3GkDk6Fl6+Mh6F99ZYO4559vMf0VK6zKbyjuoTtO4fBB0Xixfj307cuSr9cD0O73ZxW5mIMJ+i+/2GIK0Vi61FIWzzgj8vFIqYvBhS0KKuib8qmxGVz9LlhvO1LYZeNGy/bxxRQcp2C4oMeDxYutOMrSpSy59SXg4AHGoqJtW0tFDK7vGIl58+w5WKc7nKCgh9bHDK4jWhBB//57azN+fOTje/eaV37BBWaLSHRBr1+fQy5C5jhHGi7oh8oXX0Dv3qZW33zD4rKdqFQptxAWJaGZLtGYN89CLeGhjSCtW1sWTKh3HZqDHqRBA5s0tW9f5H5mzrQvhd//3sI84Xz+uZWTHTTI0i5btbIFlsJJ9Gn/jlNcuKAXlowMm+35f/9ntWd//BG6d2fJEhPZaAtGxJtYBH3uXEuNDI+fB4mU6bJmjZUaCC2rGxTZaJOLli0zr3rxYluLI5wRI2wQ9pRTbLt9+7w9dMdxCoYLemFYtcrSRZ5+2tYTmzbN3E1MzKJ5wkVBjRomtPl56J07Rz8eTdCbNjVRD5Lf5KJly2zSUt++8MgjtohzkPR0+OwzuPjinKyWDh3M7oyM3P24oDtO4YhJ0EWkv4gsFZEVIjIswvHnRWRO4LFMRLbF39QSwkcfWQB40SJb0/PVV7PLD+7ebbHnwxU/D3LMMdFz0bdsgbVr8xb04NKj4YIeHjaKRdCPOcZWvN+8OXeZ3zFjLA/+0ktz9nXoYGK+YkXOPlXr3wXdcQpOvoIuIknAy8CZQHtgkIi0D22jqneparKqJgMvAR8XhbHFzlNP2TqebdtarmDYNMdly0yQDqeHDiai0Tz04IBo6Orz4VSqZKGVUEEPnSUaJC9B37vXvgTatoVu3azGzPPP59RaHznSPP7eIetadehgz6Fhlx07LGPHY+iOU3Bi8dB7AitUdaWq7gdGAufm0X4QMCIexpUoPvgA7r/fRvSmTMmJU4SwZIk9F4eHvnmz5Y2HExT0vDx0yJ2LfuCAzWCNJuiRYug//2xfZm3b2vYTT1i45v77za4JE2DgwNxjC+3aHZzp4jnojlN4YhH0JsCvIdupgX0HISItgFbAV1GODxGRGSIyY1N+CcsliRkzbPn2E07IsyTikiU5szcPJ3kNjM6da+KYn8cbmou+fr1lqbRokbtN5cq2sEYkDz0Y8gnee5MmcO+9FpW67z4LrQwadHB/rVvnznRxQXecwhOLoEfKBtYI+wAuBT5U1QORDqrqa6qaoqop9RJl9d/UVBgwwBRm9Og8i64sXmye7mGYT5SLvAR93ry8wy1BWre2W923L3LKYpBouejhgg4m6I0awX/+Y6vdd+t28HnhmS4+7d9xCk8sgp4KNAvZbgqkRWl7KaUp3LJrl4l5errNnc/HbVyy5PDHz8G+RMqWPVjQMzNNLPMLt4AJuqrFzvMS9Pr1owt6w4a5Z3dWrWqhFzDvPNJEoQ4d7Nxgpot76I5TeGKp5TIdaCMirYC1mGhfFt5IRI4BagE/xtXC4iIry0b25s61FI1OnfJsfuCACWq06fVFSdmycNRRBwv68uU2WBmroIOFXYKzRJs1O7hdgwbWbzjLluXEz0O56ip7by68MPJ1g5kuy5ebtx4U9Lp187fZcZzc5Ouhq2omMBSYACwGRqnqQhF5TEQGhDQdBIxU1WjhmMTiz3+Gjz+GZ5+11RjyYfVqC1cUh4cOkTNdYslwCRIq6GvWWGXGSLVU8gq5RBL0pCRbKq9WrcjXDc902bjR2saycpPjOLmJqdqiqo4DxoXtezhs+9H4mVXMLFligj5oENx5Z0ynBFcpKk5BHz/evOHgZKB588x7j8Wmhg0t9h8U9PAB0SANGlhue2jZ2+3bTYgjCXp+tGtnA8kLF9qkI5/27ziFx2eKhqMKt95qRcKffz7mClHBlMXiFPT9+3PCJWDRonbtYls8o0yZnNTFSJOKgjRoYG9RaJJSMARTGEGvVMl+HYR66B4/d5zC4YIezsiRVrT7yScPchV/+gnuvtvC6+EEl52rU+cw2RlGpEyXWDNcgrRqZROB8hN0yB12iZThUhDat89JXXRBd5zC44IeyvbtViowJcVqtITx3//atPaxYw8+dfHiwz+hKJRwQf/tN6tpHsuAaJDWre2LaevWggu6iA3MFoZgpsv+/T7t33EOBRf0UB55xBTllVdyV6UKEJx48/jjuWuHQ/GlLAapW9cGE4OCPn++PRdU0PfutdcFFfSWLQu3LiqYoGdmmpe+davH0B2nsLigB5kzB156CW6+2Tz0CKxcabnVM2dabe8gmzbZQGFxeugiuTNdYp3yH0poNYO8BkXhYEEvTPw8SDDT5euv7dk9dMcpHC7oYEHxm2+2APif/xy1yS+/WApeixa5vfTiHhANEiroc+ea196oUeznhwp6NA+9WjXLhgnmi6seuqAHM10mT7ZtF3THKRwu6GD1WaZOtZzzKAnT69bl5JkPG2bNJ02yY8GUxeL00MEEPS3NJrYGa6AXZBm3QEl3ypa1NMZIiOTORd+wwa53KIJesaLF37/5xrY95OI4hcMFPT3dSgKeeCJceWXUZsE1O1u3hmuvteJTjz1m+5YssUJTkWZWHk6CorpkCSxYULBwC1g4qX79gxe2CCdU0IMZLoci6GCZLsEFMdxDd5zC4YL+/PMWBH/uuTzd2eCAaOvWNvh3333w3XfmVR7uZeeiEcx0GTfOFpMoSMpikGOPzV+cIwn6oVaYDMbRwQXdcQrLkS3omzZZmOXCC6FHjzybrlxpgh2MLd9wgwnb448f/mXnonH00fadNGqUbRfUQwd4912LQOVFqKAvX27T9A91UeygoJcvH7nkgOM4+XNkC/qTT1pFxSgDoaGsXGmiFVxouVIlKw87aZItMVrc8XOwWHTLlpb+l5RkYYyC0qRJ7oWhI9GggX0XZmWZh3700XmHaGIhKOgNGhQs7u84Tg6lVtAzMmwGf2pqlAarV8M//2kB8Rjc65UrD16k6KabcqoClgQPHXLCLsccU3R12Rs0sJoxW7YceoZLkGDIysMtjlN4Sq2gL19ueh1pVicAjz5qruAjj8TUXyRBr1LFSgFA4cIbRUFQ0IvSnqDopqXZAs/xEPSKFc32JhHXwnIcJxZiqraYiKSn2/O6dREOLlwIb78Nd90VU2rKrl0WM46wjCj33gunnJIjpMXN4RD0YFrh9Ok2XT8egg62XF3lyvHpy3GOREqtoO/cac9pkdZWevBBy9G7//6Y+gquXB9J0JOSoGfPwtlYFASFPMpk17gQFPTvvrPneAl6PmuIOI6TD6VW0KN66FOnwiefWHpKjKURQ3PQSzonnGBVIfNJ2jkkwgX9cC+K7ThOZGKKoYtIfxFZKiIrRGRYlDaXiMgiEVkoIu/F18yCE9VD/+MfLQgc48IVkDsHvaQjYr8YijJTpFYty/b55RcrBeAzOx2nZJCvhy4iScDLwOnYgtHTRWSMqi4KadMGuB/orapbRaTYcxUieugrV1rBkKeftpBLjKxcabnRtWvH18ZERcS+E9eutXCLpxk6TskgFg+9J7BCVVeq6n5gJHBuWJsbgJdVdSuAqm6Mr5kFJ+ihb9yYs6I8n35qzxddVKC+Vq60WiMuXDkEvfJ4xc8dxzl0YhH0JsCvIdupgX2htAXaisj3IjJVRPpH6khEhojIDBGZsSl0DbMiIOihq4aUev30Uxt5K2DsJFLK4pGOC7rjlDxiEfRIfmnY8g6UBdoA/YBBwOsiUvOgk1RfU9UUVU2pV69eQW0tEEEPHQJhl82bbRTvvPMK1E+wbK4Lem5c0B2n5BFLlksqEJqs3RQIH2pMBaaqagbwi4gsxQR+elysLARBDx0CA6OL/mfqfG54tChvgmVzXdBz44LuOCWPWDz06UAbEWklIuWBS4ExYW0+AU4GEJG6WAhmZTwNLSg7d+aMe65bh6UqNm0K3boVqJ9EynA5nLRrZ++vC7rjlBzyFXRVzQSGAhOAxcAoVV0oIo+JyIBAswnAFhFZBEwG7lXVLUVldCykp9tAZpkykLY6A774wrzzAo5sJlIO+uHkyiutKJlXRnSckkNME4tUdRwwLmzfwyGvFfh94FEi2LkTata09Lq0mWmwe3eBwy1wcNlcx0hKinleluM4h4lSW5wrPd1CAo0bw7rF282V7Nu3wP2sXGnlXsqXLwIjHcdx4kipnfq/c6fNYhSU1HkCF59dKFX2lEXHcRKFUu+hNyqzgXWZ9QoVbgEXdMdxEodSK+hBD73x5nlspD6Zp59Z4D7yKpvrOI5T0iiVgq4aSFusojRa8R1KGTbsKXg6Rl5lcx3HcUoapVLQd+82Ua+2ZyONN80BotRFzwfPQXccJ5EolYIenCVadflsGmHlFl3QHccp7ZRKQQ/Wcam2cCqNu1gl34hL0eXDzz9btqPnWzuOkwiUSkHP9tB/nkP9C/rYbNFCeuitW3vZXMdxEoNSKejZHjrplD39ZOrXL5yH7imLjuMkEqVS0LM99KS9kJxM48YF99C9bK7jOIlGqRT0bA+9TUOoVIlGjQou6F4213GcRKNUCnr6Dlt/o2pXW46+ceOCh1w8w8VxnESjVAr6zlWbAajW81gAGjWytUUzM2PvwwXdcZxEo1QKevoyc8er9u4CmIeea23RGFi50rJbWrQoCgsdx3HiT6kU9J2/bKICeymX3AEwQYeCxdGXL7cFjrxsruM4iUJMgi4i/UVkqYisEJFhEY5fIyKbRGRO4DE4/qbGTvra7VQtuxfKlQMs5AKxx9FV4dtvoVevIjLQcRynCMi3HrqIJAEvA6dji0FPF5ExqroorOn7qjq0CGwsGJmZ7Ny4h2pVDmTvKqiHvnQprF0Lp51WBPY5juMUEbF46D2BFaq6UlX3AyOBwhUXPxwsXkz6gUpUrZ5za/XrWzw8VkGfONGeTz+9COxzHMcpImIR9CbAryHbqYF94VwoIvNE5EMRaRapIxEZIiIzRGTGpk2bCmFuDEyfzk6qUq1uxexdZctCgwaxh1wmToRWrTzDxXGcxCIWQY9UyUTDtscCLVW1MzAReCtSR6r6mqqmqGpKvXr1CmZprEybRnpSTaqGCDoQ82zRzEyYPNnDLY7jJB6xCHoqEOpxNwVySaOqblHVfYHNfwPd42NeIZg+nZ0V61Gteu7voUaNYvPQZ8yAHTtc0B3HSTxiEfTpQBsRaSUi5YFLgTGhDUSkUcjmAGBx/EwsAHv3wrx55qFXzX0oVg89GD8/5ZT4m+c4jlOU5JvloqqZIjIUmAAkAW+o6kIReQyYoapjgNtFZACQCfwGXFOENkdn7lzLctHKVKuW+1DobNGyedz1xInQtSvUrVu0pjqO48SbfAUdQFXHAePC9j0c8vp+4P74mlYIpk8HIH1f+YgeenC2aJNIQ7rYotA//AB33lnEdjqO4xQBpWum6PTp7K/flP37JaKHDnmHXb77DjIyPH7uOE5iUuoEfWdyH4CIHjrkPTA6caJN9e/Tp4jscxzHKUJKj6Cnp8OSJezsYPP1wz30WGaLTpwIvXtD5cpFZKPjOE4RUnoEfeZMUCW9rWVMhnvowdmi0Tz0jRttTNXDLY7jJCqlR9ADA6I7W3UCDvbQg7NFo3noX31lzy7ojuMkKqVL0Fu2JD2pJnCwhw7kuRTdxIlQowZ0L74pUY7jOIdE6RL0Hj1y1hOtdnCTaEvRqcKXX9pkoqSkojXTcRynqCgdgr5pE6xaBT16kJ5uuwriof/8M6xZ4+EWx3ESm9Ih6LNn23NKSr4eeqS1RYPT/V3QHcdJZEqHoK9ZY89HHZWnhx5pbdEDB+DTT6FZM2jTpuhNdRzHKSpKh6CvXWvPDRuyc6elJ0bKJQ9fiu7nn+Hkk2H8eLj6ajvPcRwnUYmplkuJJy3NEs3Llyc93bzzSOIcnFy0dq2Vyb3nHhsEffNNuOqqw2qx4zhO3Ck9gh5Q6507I8fPIcdDv/FGC7ucfjr85z8WbnEcx0l0SkfIJUTQgx56JBo0sFotO3fCK6/AhAku5o7jlB5Kh4e+di106wbk7aGXLWszQps0gZYtD595juM4h4PEF/SMDMtFDBQ5z8tDByu+5TiOUxqJKeQiIv1FZKmIrBCRYXm0u0hEVERS4mdiPmzYYLmIMcTQHcdxSjP5CrqIJAEvA2cC7YFBItI+QrtqwO3AT/E2Mk+CKYsxxNAdx3FKM7F46D2BFaq6UlX3AyOBcyO0exz4K7A3jvblT3AufyDk4h664zhHKrEIehPg15Dt1MC+bESkK9BMVT/LqyMRGSIiM0RkxqZNmwpsbESCgu4euuM4RzixCHqk+ZOafVCkDPA8cHd+Hanqa6qaoqop9erVi93KvEhLs9lB9eqRlWULPbuH7jjOkUgsgp4KhGZrNwVCaxZWAzoCX4vIKuA4YMxhGxhdu9ZmDJUpw+7dNj7qHrrjOEcisQj6dKCNiLQSkfLApcCY4EFV3a6qdVW1paq2BKYCA1R1RpFYHE5aWq74ObiH7jjOkUm+gq6qmcBQYAKwGBilqgtF5DERGVDUBuZL2CxRcA/dcZwjk5gmFqnqOGBc2L6Ho7Ttd+hmFYC1a6GfXdI9dMdxjmQSu5bL7t2wbVuuWaLgHrrjOEcmiS3owcLmIbNEwT10x3GOTBJb0CPMEgX30B3HOTJJbEEPm1TkHrrjOEcypUPQPYbuOI5TCgS9UiWoUQPI8dBd0B3HORJJbEFfu9bCLYEFRNPToUIFKFeumO1yHMcpBhJb0ENmiYJXWnQc58gm8QU9MCAKXmnRcZwjm8QVdNWckEsA99AdxzmSSVxB374d9uzJFXJxD91xnCOZxBX0sBx0cA/dcZwjm1Il6O6hO45zJJO4gh427R/cQ3cc58gmcQXdPXTHcZxcJLag16wJlStn73IP3XGcI5mYBF1E+ovIUhFZISLDIhy/SUTmi8gcEZkiIu3jb2oYYSmL+/fbwz10x3GOVPIVdBFJAl4GzgTaA4MiCPZ7qtpJVZOBvwJ/i7ul4USYJQruoTuOc+QSi4feE1ihqitVdT8wEjg3tIGq7gjZrAJo/EyMQoRZouAeuuM4Ry6xrCnaBPg1ZDsV6BXeSERuBX4PlAdOidSRiAwBhgA0b968oLbmkJVlqxWFZbiAe+iO4xy5xOKhS4R9B3ngqvqyqh4F/AF4MFJHqvqaqqaoakq9evUKZmkomzZBZqZ76I7jOCHEIuipQLOQ7aZAWh7tRwLnHYpR+RK2sAW4h+44jhOLoE8H2ohIKxEpD1wKjAltICJtQjbPBpbHz8QIRMlBB/fQHcc5csk3hq6qmSIyFJgAJAFvqOpCEXkMmKGqY4ChInIakAFsBa4uSqOjzRIF99AdxzlyiWVQFFUdB4wL2/dwyOs74mxX3qSl2SpFDRtm73IP3XGcI53EnCmalgb16+daa849dMdxjnQSV9BDwi1gHnqZMrZmtOM4zpFIYgp62LR/MA+9atXs9aIdx3GOOBJT0MOm/YNXWnQcx0k8Qc/IgI0bI3roHj93HOdIJvEEfd06e44QQ3cP3XGcI5nEE/QIs0TBPXTHcZzEFXT30B3HcXIR08SiEkWEWaLgHrqTuGRkZJCamsrevXuL2xSnBFGxYkWaNm1KuZD5NvmReILeogVcdBHUrZtrt3voTqKSmppKtWrVaNmyJeJ5tw6gqmzZsoXU1FRatWoV83mJF3IZMAA++MBmEYXgHrqTqOzdu5c6deq4mDvZiAh16tQp8K+2xBP0CGRl5UwscpxExMXcCacwfxOlQtB377Zn99AdxzmSKRWC7pUWHafwbNmyheTkZJKTk2nYsCFNmjTJ3t6/f39MfVx77bUsXbo0zzYvv/wy7777bjxMdqKQeIOiEfBKi45TeOrUqcOcOXMAePTRR6latSr33HNPrjaqiqpSpkxkH3D48OH5XufWW289dGMPM5mZmZQtmzgyGZOHLiL9RWSpiKwQkWERjv9eRBaJyDwRmSQiLeJvanTcQ3dKDXfeCf36xfdx552FMmXFihV07NiRm266iW7durFu3TqGDBlCSkoKHTp04LHHHstu26dPH+bMmUNmZiY1a9Zk2LBhdOnSheOPP56NGzcC8OCDD/LCCy9ktx82bBg9e/bkmGOO4YcffgBg165dXHjhhXTp0oVBgwaRkpKS/WUTyiOPPEKPHj2y7VO1ZY6XLVvGKaecQpcuXejWrRurVq0C4Mknn6RTp0506dKFBx54IJfNAOvXr+foo48G4PXXX+fSSy/lnHPO4cwzz2THjh2ccsopdOvWjc6dO/PZZ59l2zF8+HA6d+5Mly5duPbaa9m2bRutW7cmMzMTgG3bttGqVSsOHDhQqM+goOQr6CKSBLwMnAm0BwaJSPuwZrOBFFXtDHwI/DXehuaFe+iOUzQsWrSI66+/ntmzZ9OkSROeeuopZsyYwdy5c/nyyy9ZtGjRQeds376dvn37MnfuXI4//njeeOONiH2rKtOmTeOZZ57J/nJ46aWXaNiwIXPnzmXYsGHMnj074rl33HEH06dPZ/78+Wzfvp3x48cDMGjQIO666y7mzp3LDz/8QP369Rk7diyff/4506ZNY+7cudx999353vePP/7IO++8w5dffkmlSpX49NNPmTVrFhMnTuSuu+4CYO7cuTz99NN8/fXXzJ07l+eee46aNWvSu3fvbHvee+89LrnkEpKSkvJ/s+NALL8legIrVHUlgIiMBM4Fsj9JVZ0c0n4qcEU8jcyPLVvs2QXdSXgCHmxJ4aijjqJHjx7Z2yNGjOA///kPmZmZpKWlsWjRItq3z+3fVapUiTPPPBOA7t27891330Xs+4ILLshuE/Skp0yZwh/+8AcAunTpQocOHSKeO2nSJJ555hn27t3L5s2b6d69O8cddxybN2/md7/7HWATcwAmTpzIddddR6XAYgm1a9fO977POOMMatWqBdgXzx/+8AemTJlCmTJl+PXXX9m8eTNfffUVAwcOzO4v+Dx48GBefPFFzjnnHIYPH84777yT7/XiRSyC3gT4NWQ7FeiVR/vrgc8PxaiC8s03ULEidOx4OK/qOKWfKlWqZL9evnw5f//735k2bRo1a9bkiiuuiJgnXb58+ezXSUlJ2eGHcCpUqHBQm2DoJC92797N0KFDmTVrFk2aNOHBBx/MtiNSqp+qRtxftmxZsrKyAA66j9D7fvvtt9m+fTuzZs2ibNmyNG3alL1790btt2/fvgwdOpTJkydTrlw52rVrl+89xYtYYuiRkiEjvusicgWQAjwT5fgQEZkhIjM2bdoUu5X5MH68hQp9tSLHKTp27NhBtWrVqF69OuvWrWPChAlxv0afPn0YNWoUAPPnz48Y0tmzZw9lypShbt26pKen89FHHwFQq1Yt6taty9ixYwET6d27d3PGGWfwn//8hz179gDw22+/AdCyZUtmzpwJwIcffhjVpu3bt1O/fn3Kli3Ll19+ydpA+ZHTTjuNkSNHZvcXfAa44ooruPzyy7n22msP6f0oKLEIeirQLGS7KZAW3khETgMeAAao6r5IHanqa6qaoqop9erVK4y9B/HLL7B0KfTvH5fuHMeJQrdu3Wjfvj0dO3bkhhtuoHfv3nG/xm233cbatWvp3Lkzzz33HB07dqRGjRq52tSpU4err76ajh07cv7559OrV07A4N133+W5556jc+fO9OnTh02bNnHOOefQv39/UlJSSE5O5vnnnwfg3nvv5e9//zsnnHACW7dujWrTlVdeyQ8//EBKSgoffPABbdq0AaBz587cd999nHTSSSQnJ3Pvvfdmn3P55Zezfft2Bg4cGM+3J3+C6UjRHlhYZiXQCigPzAU6hLXpCvwMtMmvv+Cje/fuGg/++U9VUF26NC7dOc5hZ9GiRcVtQokhIyND9+zZo6qqy5Yt05YtW2pGRkYxW1VwRowYoddcc80h9xPpbwOYoVF0Nd8YuqpmishQYAKQBLyhqgtF5LFAx2OwEEtV4INATGmNqg6I83dPRMaPh1atIPCl6ThOArNz505OPfVUMjMzUVVeffXVhMoDB7j55puZOHFidqbL4SSmd0pVxwHjwvY9HPL63TwerQAAC/NJREFUtDjbFRP79sGkSXDVVb44tOOUBmrWrJkd105UXnnllWK7dkJP/f/+e9i1CwIZUo7jOEc0CS3on38O5cvDyScXtyWO4zjFT0IL+vjxcOKJPuXfcRwHEljQf/0VFizwdEXHcZwgCSvowTkNLuiOc2j069fvoElCL7zwArfcckue51UN/DROS0vjoosuitr3jBkz8uznhRdeYHdwUQPgrLPOYtu2bbGY7oSRsII+fjw0bQpRSj04jhMjgwYNYuTIkbn2jRw5kkGDBsV0fuPGjfOcaZkf4YI+btw4atasWej+Djeqml1CoLhJSEHPyIAvvzTv3NMVndJEcVTPveiii/jss8/Yt88meK9atYq0tDT69OmTnRferVs3OnXqxKeffnrQ+atWraJjoJDSnj17uPTSS+ncuTMDBw7Mnm4Plp8dLL37yCOPAPDiiy+SlpbGySefzMmB7IaWLVuyefNmAP72t7/RsWNHOnbsmF16d9WqVRx77LHccMMNdOjQgTPOOCPXdYKMHTuWXr160bVrV0477TQ2bNgAWK77tddeS6dOnejcuXN26YDx48fTrVs3unTpwqmnngpYffhnn302u8+OHTuyatWqbBtuueUWunXrxq+//hrx/gCmT5/OCSecQJcuXejZsyfp6emceOKJucoC9+7dm3nz5uX9QcVAYmXsB5g6FXbs8HCL48SDOnXq0LNnT8aPH8+5557LyJEjGThwICJCxYoVGT16NNWrV2fz5s0cd9xxDBgwIOp6l6+88gqVK1dm3rx5zJs3j27dumUfe+KJJ6hduzYHDhzg1FNPZd68edx+++387W9/Y/LkydStWzdXXzNnzmT48OH89NNPqCq9evWib9++1KpVi+XLlzNixAj+/e9/c8kll/DRRx9xxRW5i7z26dOHqVOnIiK8/vrr/PWvf+W5557j8ccfp0aNGsyfPx+ArVu3smnTJm644Qa+/fZbWrVqlasuSzSWLl3K8OHD+ec//xn1/tq1a8fAgQN5//336dGjBzt27KBSpUoMHjyYN998kxdeeIFly5axb98+OnfuXKDPLRIJKeiffw5JSXBasUxncpyio7iq5wbDLkFBD9YwV1X++Mc/8u2331KmTBnWrl3Lhg0baNiwYcR+vv32W26//XbAap2EitSoUaN47bXXyMzMZN26dSxatChPEZsyZQrnn39+duXDCy64gO+++44BAwbQqlUrkpOTgdzld0NJTU1l4MCBrFu3jv3799OqVSvAyumGhphq1arF2LFjOemkk7LbxFJit0WLFhx33HF53p+I0KhRo+wSxNWrVwfg4osv5vHHH+eZZ57hjTfe4Jprrsn3erGQkCGX8ePhhBMgrGaP4ziF5LzzzmPSpEnMmjWLPXv2ZHvW7777Lps2bWLmzJnMmTOHBg0aRCyZG0ok7/2XX37h2WefZdKkScybN4+zzz473340j1K6wdK7EL1E72233cbQoUOZP38+r776avb1NELZ20j7IHeJXchdZje0xG60+4vWb+XKlTn99NP59NNPGTVqFJdddlnUey0ICSfo69fD7Nk+O9Rx4knVqlXp168f1113Xa7B0GDp2HLlyjF58mRWr16dZz8nnXRS9kLQCxYsyI4L79ixgypVqlCjRg02bNjA55/nLJlQrVo10oPrSIb19cknn7B792527drF6NGjOfHEE2O+p+3bt9OkSRMA3nrrrez9Z5xxBv/4xz+yt7du3crxxx/PN998wy+//ALkLrE7a9YsAGbNmpV9PJxo99euXTvS0tKYPn06AOnp6dlfPoMHD+b222+nR48eMf0iiIWEE3RPV3ScomHQoEHMnTuXSy+9NHvf5ZdfzowZM0hJSeHdd9/Nd7GGm2++mZ07d9K5c2f++te/0rNnT8BWH+ratSsdOnTguuuuy1V6d8iQIZx55pnZg6JBunXrxjXXXEPPnj3p1asXgwcPpmvXrjHfz6OPPsrFF1/MiSeemCs+/+CDD7J161Y6duxIly5dmDx5MvXq1eO1117jggsuoEuXLtllby+88EJ+++03kpOTeeWVV2jbtm3Ea0W7v/Lly/P+++9z22230aVLF04//fRsL7979+5Ur149rjXTJa+fNUVJSkqK5pefGokxY+CNN+DjjyHKAuSOk1AsXryYY489trjNcA4zaWlp9OvXjyVLllAmiphF+tsQkZmqmhKpfcJJ4oAB8MknLuaO4yQub7/9Nr169eKJJ56IKuaFISGzXBzHcRKZq666iquuuiru/bqf6zglgOIKfToll8L8TcQk6CLSX0SWisgKERkW4fhJIjJLRDJFJHJRB8dxIlKxYkW2bNniou5ko6ps2bKFihUrFui8fEMuIpIEvAycji0YPV1Exqhq6HLca4BrgHsKdHXHcWjatCmpqals2rSpuE1xShAVK1akadOmBTonlhh6T2CFqq4EEJGRwLlAtqCr6qrAsZJRocZxEohy5cplz1B0nEMhlpBLE+DXkO3UwD7HcRynBBGLoEeqwlOoYJ+IDBGRGSIyw39eOo7jxJdYBD0VaBay3RRIK8zFVPU1VU1R1ZR69eoVpgvHcRwnCrHE0KcDbUSkFbAWuBQ45EoyM2fO3CwieReGiE5dYPOh2lCCKE33U5ruBfx+SjKl6V4g9vtpEe1ATFP/ReQs4AUgCXhDVZ8QkceAGao6RkR6AKOBWsBeYL2qFtlaQiIyI9rU10SkNN1PaboX8PspyZSme4H43E9MM0VVdRwwLmzfwyGvp2OhGMdxHKeY8JmijuM4pYREFfTXituAOFOa7qc03Qv4/ZRkStO9QBzup9jK5zqO4zjxJVE9dMdxHCcMF3THcZxSQsIJen6VH0s6IvKGiGwUkQUh+2qLyJcisjzwXKs4bYwVEWkmIpNFZLGILBSROwL7E/V+KorINBGZG7ifPwX2txKRnwL3876IlC9uW2NFRJJEZLaIfBbYTuR7WSUi80VkjojMCOxL1L+1miLyoYgsCfz/HB+Pe0koQQ+p/Hgm0B4YJCLti9eqAvMmEL4i6jBgkqq2ASYFthOBTOBuVT0WOA64NfB5JOr97ANOUdUuQDLQX0SOA54Gng/cz1bg+mK0saDcASwO2U7kewE4WVWTQ/K1E/Vv7e/AeFVtB3TBPqNDvxdVTZgHcDwwIWT7fuD+4rarEPfRElgQsr0UaBR43QhYWtw2FvK+PsXKLCf8/QCVgVlAL2z2XtnA/lx/gyX5gc0NmQScAnyG1WVKyHsJ2LsKqBu2L+H+1oDqwC8EklLieS8J5aFTeis/NlDVdQCB5/rFbE+BEZGWQFfgJxL4fgIhijnARuBL4Gdgm6pmBpok0t/cC8B9QLCsdR0S917AigJ+ISIzRWRIYF8i/q21BjYBwwPhsNdFpApxuJdEE/S4VX504oeIVAU+Au7U/2/vDlqiisIwjv9fKCJCEKGF0CLaiJswF26MCIoWEq4DFy5a9gkk8COEO1ctpUCJELel26IoxRCshaBUzioCVxGPi3NuzEJBZ4bunMvzg8s9c+Yu3hfOfWfmHOZc6Xfd8XRD0l9JY6RvtxPA6EmX/d+ozi8iHgItSR/bu0+4tO9zaTMpaZw05fokIu7UHVCHLgDjwKKkW8ARPZoqKq2g92znxz5zGBHDAPncqjmeM4uIi6RiviTpVe4uNp+KpF/ABmltYDAiqm0yShlzk8B0ROwBL0nTLguUmQsAkr7nc4u0d9QEZY61A+BA0rv8eoVU4LvOpbSC/m/nx7w6/whYrTmmXlgFZnN7ljQX3fciIoDnwI6kZ21vlZrP1YgYzO3LwH3SYtU6UD0rt4h8JM1JuibpOuk+eStphgJzAYiIKxExULWBB8A2BY41ST+B/YgYyV33SE+A6z6XuhcIOlhQmAJ2SXObT+uOp4P4XwA/gD+kT+rHpLnNN8DXfB6qO84z5nKb9JN9C/icj6mC87kJfMr5bAPzuf8G8B74BiwDl+qO9Zx53QXWSs4lx72Zjy/VvV/wWBsDPuSx9pq0U23Xufiv/2ZmDVHalIuZmZ3CBd3MrCFc0M3MGsIF3cysIVzQzcwawgXdzKwhXNDNzBriGCTNA8/6KAMJAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "acc = history.history['accuracy']\n",
    "val_acc = history.history['val_accuracy']\n",
    "loss = history.history['loss']\n",
    "val_loss = history.history['val_loss']\n",
    "\n",
    "epochs = range(len(acc))\n",
    "\n",
    "plt.plot(epochs, acc, 'r', label='Training accuracy')\n",
    "plt.plot(epochs, val_acc, 'b', label='Validation accuracy')\n",
    "plt.title('Training and validation accuracy')\n",
    "plt.legend()\n",
    "plt.figure()\n",
    "\n",
    "plt.plot(epochs, loss, 'r', label='Training Loss')\n",
    "plt.plot(epochs, val_loss, 'b', label='Validation Loss')\n",
    "plt.title('Training and validation loss')\n",
    "plt.legend()\n",
    "\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "test accuracy: 97.28109240531921\n"
     ]
    }
   ],
   "source": [
    "print(\"test accuracy: \"+ str(model.evaluate_generator(testdata_generator)[1]*100))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 61/100\n",
      "687/687 [==============================] - 8s 12ms/step - loss: 0.3184 - accuracy: 0.8958 - val_loss: 0.1314 - val_accuracy: 0.9055\n",
      "Epoch 62/100\n",
      "687/687 [==============================] - 8s 12ms/step - loss: 0.3008 - accuracy: 0.8973 - val_loss: 0.1739 - val_accuracy: 0.9317\n",
      "Epoch 63/100\n",
      "687/687 [==============================] - 8s 12ms/step - loss: 0.3011 - accuracy: 0.8987 - val_loss: 0.4440 - val_accuracy: 0.8304\n",
      "Epoch 64/100\n",
      "687/687 [==============================] - 8s 12ms/step - loss: 0.3145 - accuracy: 0.8969 - val_loss: 0.3446 - val_accuracy: 0.9179\n",
      "Epoch 65/100\n",
      "687/687 [==============================] - 8s 12ms/step - loss: 0.2986 - accuracy: 0.8992 - val_loss: 0.1552 - val_accuracy: 0.9073\n",
      "Epoch 66/100\n",
      "687/687 [==============================] - 8s 12ms/step - loss: 0.3025 - accuracy: 0.8996 - val_loss: 0.1296 - val_accuracy: 0.8840\n",
      "Epoch 67/100\n",
      "687/687 [==============================] - 8s 12ms/step - loss: 0.2895 - accuracy: 0.9023 - val_loss: 0.0574 - val_accuracy: 0.8884\n",
      "Epoch 68/100\n",
      "687/687 [==============================] - 8s 12ms/step - loss: 0.3016 - accuracy: 0.9006 - val_loss: 0.1773 - val_accuracy: 0.9288\n",
      "Epoch 69/100\n",
      "687/687 [==============================] - 8s 12ms/step - loss: 0.2965 - accuracy: 0.8992 - val_loss: 0.0577 - val_accuracy: 0.8740\n",
      "Epoch 70/100\n",
      "687/687 [==============================] - 8s 12ms/step - loss: 0.2759 - accuracy: 0.9091 - val_loss: 0.3796 - val_accuracy: 0.8170\n",
      "Epoch 71/100\n",
      "687/687 [==============================] - 8s 12ms/step - loss: 0.2981 - accuracy: 0.8992 - val_loss: 0.3522 - val_accuracy: 0.9168\n",
      "Epoch 72/100\n",
      "687/687 [==============================] - 8s 12ms/step - loss: 0.2867 - accuracy: 0.9023 - val_loss: 0.6836 - val_accuracy: 0.8263\n",
      "Epoch 73/100\n",
      "687/687 [==============================] - 8s 12ms/step - loss: 0.2869 - accuracy: 0.9041 - val_loss: 0.1443 - val_accuracy: 0.9377\n",
      "Epoch 74/100\n",
      "687/687 [==============================] - 8s 12ms/step - loss: 0.2742 - accuracy: 0.9087 - val_loss: 0.1181 - val_accuracy: 0.9434\n",
      "Epoch 75/100\n",
      "687/687 [==============================] - 8s 12ms/step - loss: 0.2653 - accuracy: 0.9096 - val_loss: 0.4208 - val_accuracy: 0.9049\n",
      "Epoch 76/100\n",
      "687/687 [==============================] - 8s 12ms/step - loss: 0.2722 - accuracy: 0.9084 - val_loss: 0.1920 - val_accuracy: 0.9381\n",
      "Epoch 77/100\n",
      "687/687 [==============================] - 8s 12ms/step - loss: 0.2762 - accuracy: 0.9088 - val_loss: 0.3779 - val_accuracy: 0.9423\n",
      "Epoch 78/100\n",
      "687/687 [==============================] - 8s 12ms/step - loss: 0.2850 - accuracy: 0.9051 - val_loss: 0.1172 - val_accuracy: 0.9160\n",
      "Epoch 79/100\n",
      "687/687 [==============================] - 8s 12ms/step - loss: 0.2772 - accuracy: 0.9073 - val_loss: 0.0500 - val_accuracy: 0.9466\n",
      "Epoch 80/100\n",
      "687/687 [==============================] - 8s 12ms/step - loss: 0.2773 - accuracy: 0.9065 - val_loss: 0.3109 - val_accuracy: 0.9348\n",
      "Epoch 81/100\n",
      "687/687 [==============================] - 8s 12ms/step - loss: 0.2626 - accuracy: 0.9110 - val_loss: 0.2869 - val_accuracy: 0.9057\n",
      "Epoch 82/100\n",
      "687/687 [==============================] - 8s 12ms/step - loss: 0.2675 - accuracy: 0.9094 - val_loss: 0.2440 - val_accuracy: 0.9310\n",
      "Epoch 83/100\n",
      "687/687 [==============================] - 8s 12ms/step - loss: 0.2659 - accuracy: 0.9130 - val_loss: 0.8335 - val_accuracy: 0.7725\n",
      "Epoch 84/100\n",
      "687/687 [==============================] - 8s 12ms/step - loss: 0.2650 - accuracy: 0.9119 - val_loss: 0.1309 - val_accuracy: 0.9217\n",
      "Epoch 85/100\n",
      "687/687 [==============================] - 8s 12ms/step - loss: 0.2759 - accuracy: 0.9067 - val_loss: 0.0176 - val_accuracy: 0.9434\n",
      "Epoch 86/100\n",
      "687/687 [==============================] - 8s 12ms/step - loss: 0.2595 - accuracy: 0.9115 - val_loss: 0.0575 - val_accuracy: 0.9221\n",
      "Epoch 87/100\n",
      "687/687 [==============================] - 8s 12ms/step - loss: 0.2630 - accuracy: 0.9111 - val_loss: 0.1837 - val_accuracy: 0.9406\n",
      "Epoch 88/100\n",
      "687/687 [==============================] - 8s 12ms/step - loss: 0.2640 - accuracy: 0.9135 - val_loss: 0.0893 - val_accuracy: 0.9204\n",
      "Epoch 89/100\n",
      "687/687 [==============================] - 8s 12ms/step - loss: 0.2514 - accuracy: 0.9178 - val_loss: 0.1226 - val_accuracy: 0.9421\n",
      "Epoch 90/100\n",
      "687/687 [==============================] - 8s 12ms/step - loss: 0.2565 - accuracy: 0.9140 - val_loss: 0.0540 - val_accuracy: 0.8962\n",
      "Epoch 91/100\n",
      "687/687 [==============================] - 8s 12ms/step - loss: 0.2588 - accuracy: 0.9137 - val_loss: 0.4131 - val_accuracy: 0.8876\n",
      "Epoch 92/100\n",
      "687/687 [==============================] - 8s 12ms/step - loss: 0.2536 - accuracy: 0.9160 - val_loss: 0.1725 - val_accuracy: 0.9434\n",
      "Epoch 93/100\n",
      "687/687 [==============================] - 9s 13ms/step - loss: 0.2589 - accuracy: 0.9147 - val_loss: 0.2241 - val_accuracy: 0.9432\n",
      "Epoch 94/100\n",
      "687/687 [==============================] - 8s 12ms/step - loss: 0.2436 - accuracy: 0.9185 - val_loss: 0.3249 - val_accuracy: 0.9441\n",
      "Epoch 95/100\n",
      "687/687 [==============================] - 8s 12ms/step - loss: 0.2646 - accuracy: 0.9105 - val_loss: 0.3807 - val_accuracy: 0.8987\n",
      "Epoch 96/100\n",
      "687/687 [==============================] - 8s 12ms/step - loss: 0.2429 - accuracy: 0.9195 - val_loss: 0.3375 - val_accuracy: 0.9432\n",
      "Epoch 97/100\n",
      "687/687 [==============================] - 9s 13ms/step - loss: 0.2533 - accuracy: 0.9164 - val_loss: 0.4208 - val_accuracy: 0.7969\n",
      "Epoch 98/100\n",
      "687/687 [==============================] - 9s 13ms/step - loss: 0.2359 - accuracy: 0.9235 - val_loss: 0.1195 - val_accuracy: 0.8955\n",
      "Epoch 99/100\n",
      "687/687 [==============================] - 8s 12ms/step - loss: 0.2510 - accuracy: 0.9170 - val_loss: 0.0601 - val_accuracy: 0.9603\n",
      "Epoch 100/100\n",
      "687/687 [==============================] - 8s 12ms/step - loss: 0.2455 - accuracy: 0.9190 - val_loss: 0.1651 - val_accuracy: 0.9335\n"
     ]
    }
   ],
   "source": [
    "history2=model.fit(traindata_generator,epochs=100,validation_data=validationdata_generator,callbacks=[callback],initial_epoch=60)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "acc = history2.history['accuracy']\n",
    "val_acc = history2.history['val_accuracy']\n",
    "loss = history2.history['loss']\n",
    "val_loss = history2.history['val_loss']\n",
    "\n",
    "epochs = range(len(acc))\n",
    "\n",
    "plt.plot(epochs, acc, 'r', label='Training accuracy')\n",
    "plt.plot(epochs, val_acc, 'b', label='Validation accuracy')\n",
    "plt.title('Training and validation accuracy')\n",
    "plt.legend()\n",
    "plt.figure()\n",
    "\n",
    "plt.plot(epochs, loss, 'r', label='Training Loss')\n",
    "plt.plot(epochs, val_loss, 'b', label='Validation Loss')\n",
    "plt.title('Training and validation loss')\n",
    "plt.legend()\n",
    "\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "test accuracy: 98.13162088394165\n"
     ]
    }
   ],
   "source": [
    "print(\"test accuracy: \"+ str(model.evaluate_generator(testdata_generator)[1]*100))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.7.4"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}