model.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 23,
   "outputs": [],
   "source": [
    "import tensorflow as tf"
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%%\n"
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "outputs": [
    {
     "data": {
      "text/plain": "'2.6.0'"
     },
     "execution_count": 24,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "tf.__version__"
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%%\n"
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "outputs": [],
   "source": [
    "class_names = ['circle', 'square', 'star', 'triangle']"
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%%\n"
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Found 11976 images belonging to 4 classes.\n",
      "Found 2994 images belonging to 4 classes.\n"
     ]
    }
   ],
   "source": [
    "from tensorflow.keras.preprocessing.image import ImageDataGenerator\n",
    "data_dir = 'shapes/'\n",
    "datagen = ImageDataGenerator(rescale=1/255.,\n",
    "                             validation_split=0.2,\n",
    "                             rotation_range=0.2,\n",
    "                             shear_range=0.2,\n",
    "                             zoom_range=0.2,\n",
    "                             width_shift_range=0.2,\n",
    "                             height_shift_range=0.2,\n",
    "                             horizontal_flip=True)\n",
    "\n",
    "#Load data in from directories and turn it into batches\n",
    "train_data = datagen.flow_from_directory(data_dir,\n",
    "                                        target_size=(200, 200),\n",
    "                                        batch_size=32,\n",
    "                                        class_mode='categorical',\n",
    "                                        subset='training')\n",
    "test_data = datagen.flow_from_directory(data_dir,\n",
    "                                             target_size=(200, 200),\n",
    "                                             batch_size=32,\n",
    "                                             class_mode='categorical',\n",
    "                                             subset='validation')"
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%%\n"
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "outputs": [],
   "source": [
    "#Create a model (AlexNet)\n",
    "model = tf.keras.Sequential([\n",
    "    tf.keras.layers.Conv2D(96, 11, strides=4, activation='relu', padding='same', input_shape=(200, 200, 3)),\n",
    "    tf.keras.layers.MaxPool2D(3, strides=2),\n",
    "    tf.keras.layers.Conv2D(256, 5, strides=4, padding='same', activation='relu'),\n",
    "    tf.keras.layers.MaxPool2D(3, strides=2),\n",
    "    tf.keras.layers.Conv2D(384, 3, strides=4, padding='same', activation='relu'),\n",
    "    tf.keras.layers.Conv2D(384, 3, strides=4, padding='same', activation='relu'),\n",
    "    tf.keras.layers.Conv2D(256, 3, strides=4, padding='same', activation='relu'),\n",
    "    tf.keras.layers.Flatten(),\n",
    "    tf.keras.layers.Dense(4096, activation='relu'),\n",
    "    tf.keras.layers.Dropout(0.5),\n",
    "    tf.keras.layers.Dense(4096, activation='relu'),\n",
    "    tf.keras.layers.Dropout(0.5),\n",
    "    tf.keras.layers.Dense(4, activation='softmax')\n",
    "])"
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%%\n"
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model: \"sequential_1\"\n",
      "_________________________________________________________________\n",
      "Layer (type)                 Output Shape              Param #   \n",
      "=================================================================\n",
      "conv2d_5 (Conv2D)            (None, 50, 50, 96)        34944     \n",
      "_________________________________________________________________\n",
      "max_pooling2d_2 (MaxPooling2 (None, 24, 24, 96)        0         \n",
      "_________________________________________________________________\n",
      "conv2d_6 (Conv2D)            (None, 6, 6, 256)         614656    \n",
      "_________________________________________________________________\n",
      "max_pooling2d_3 (MaxPooling2 (None, 2, 2, 256)         0         \n",
      "_________________________________________________________________\n",
      "conv2d_7 (Conv2D)            (None, 1, 1, 384)         885120    \n",
      "_________________________________________________________________\n",
      "conv2d_8 (Conv2D)            (None, 1, 1, 384)         1327488   \n",
      "_________________________________________________________________\n",
      "conv2d_9 (Conv2D)            (None, 1, 1, 256)         884992    \n",
      "_________________________________________________________________\n",
      "flatten_1 (Flatten)          (None, 256)               0         \n",
      "_________________________________________________________________\n",
      "dense_3 (Dense)              (None, 4096)              1052672   \n",
      "_________________________________________________________________\n",
      "dropout_2 (Dropout)          (None, 4096)              0         \n",
      "_________________________________________________________________\n",
      "dense_4 (Dense)              (None, 4096)              16781312  \n",
      "_________________________________________________________________\n",
      "dropout_3 (Dropout)          (None, 4096)              0         \n",
      "_________________________________________________________________\n",
      "dense_5 (Dense)              (None, 4)                 16388     \n",
      "=================================================================\n",
      "Total params: 21,597,572\n",
      "Trainable params: 21,597,572\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n"
     ]
    }
   ],
   "source": [
    "model.summary()"
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%%\n"
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "outputs": [],
   "source": [
    "model.compile(loss='categorical_crossentropy',\n",
    "              optimizer=tf.keras.optimizers.Adam(0.001),\n",
    "              metrics=['accuracy'])"
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%%\n"
    }
   }
  },
  {
   "cell_type": "markdown",
   "source": [],
   "metadata": {
    "collapsed": false
   }
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/5\n",
      "94/94 [==============================] - 40s 425ms/step - loss: 0.1506 - accuracy: 0.9532 - val_loss: 0.2908 - val_accuracy: 0.8764\n",
      "Epoch 2/5\n",
      "94/94 [==============================] - 39s 411ms/step - loss: 0.1053 - accuracy: 0.9666 - val_loss: 0.0932 - val_accuracy: 0.9660\n",
      "Epoch 3/5\n",
      "94/94 [==============================] - 40s 422ms/step - loss: 0.1019 - accuracy: 0.9663 - val_loss: 0.0554 - val_accuracy: 0.9851\n",
      "Epoch 4/5\n",
      "94/94 [==============================] - 40s 424ms/step - loss: 0.0596 - accuracy: 0.9846 - val_loss: 0.0242 - val_accuracy: 0.9959\n",
      "Epoch 5/5\n",
      "94/94 [==============================] - 40s 425ms/step - loss: 0.0419 - accuracy: 0.9860 - val_loss: 0.0164 - val_accuracy: 0.9932\n"
     ]
    }
   ],
   "source": [
    "history = model.fit(test_data,\n",
    "                    epochs=5,\n",
    "                    steps_per_epoch=len(test_data),\n",
    "                    validation_data=test_data,\n",
    "                    validation_steps=int(0.25*len(test_data)))"
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%%\n"
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "94/94 [==============================] - 23s 244ms/step - loss: 0.0203 - accuracy: 0.9927\n"
     ]
    },
    {
     "data": {
      "text/plain": "[0.020273853093385696, 0.9926519989967346]"
     },
     "execution_count": 32,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "model.evaluate(test_data)"
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%%\n"
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "outputs": [],
   "source": [
    "import matplotlib.pyplot as plt\n",
    "\n",
    "def plot_loss_curves(history):\n",
    "  '''\n",
    "  Returns seperate loss curves for training and validation metrics.\n",
    "  '''\n",
    "  loss = history.history['loss']\n",
    "  val_loss = history.history['val_loss']\n",
    "\n",
    "  accuracy = history.history['accuracy']\n",
    "  val_accuracy = history.history['val_accuracy']\n",
    "\n",
    "  epochs = range(len(history.history['loss']))#how many epochss did we run for?\n",
    "\n",
    "  #Plot loss\n",
    "  plt.plot(epochs, loss, label='training_loss')\n",
    "  plt.plot(epochs, val_loss, label='val_loss')\n",
    "  plt.title('loss')\n",
    "  plt.xlabel('epochs')\n",
    "  plt.legend()\n",
    "\n",
    "  #Plot accuracy\n",
    "  plt.figure()\n",
    "  plt.plot(epochs, accuracy, label='training_accuracy')\n",
    "  plt.plot(epochs, val_accuracy, label='val_accuracy')\n",
    "  plt.title('accuracy')\n",
    "  plt.xlabel('epochs')\n",
    "  plt.legend()\n"
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%%\n"
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "outputs": [
    {
     "data": {
      "text/plain": "<Figure size 432x288 with 1 Axes>",
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAXoAAAEWCAYAAABollyxAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjUuMSwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy/YYfK9AAAACXBIWXMAAAsTAAALEwEAmpwYAAAzNUlEQVR4nO3deXxU1f3/8dcnOxCWsAVISMKOrEEh7KjggoiAiiKoBav1q1L1q9ZWW+pC9Vur/fartlbbWqtYURGwoj8VRVBE1iTsIGFLQliSELYEyDrn98e9ISEkYZJMciczn+fjkQczd87MfGY073Ny7r3nijEGpZRSvivA6QKUUkrVLw16pZTycRr0Sinl4zTolVLKx2nQK6WUj9OgV0opH6dBr/yeiKSKyFVO16FUfdGgV0opH6dBr5RSPk6DXimbiISKyMsicsj+eVlEQu3H2orIZyJyQkSOicj3IhJgP/YrETkoIrkisktExjn7SZQ6X5DTBSjlRX4DDAPiAQN8AswBfgs8BmQA7ey2wwAjIr2AnwNDjDGHRCQOCGzYspWqno7olSpzOzDXGJNljMkGngXutB8rAjoCscaYImPM98ZaKKoECAX6iEiwMSbVGLPXkeqVqoIGvVJlOgFp5e6n2dsAXgL2AF+JyD4ReQLAGLMH+G/gGSBLRD4QkU4o5UU06JUqcwiILXc/xt6GMSbXGPOYMaYrMAl4tHQu3hgz3xgzyn6uAf7QsGUrVT0NeqXKvA/MEZF2ItIWeAr4N4CITBSR7iIiwEmsKRuXiPQSkbH2Ttt84Czgcqh+pSqlQa9UmeeARGALsBVItrcB9ACWAXnAGuCvxpgVWPPzLwBHgSNAe+DJhi1bqeqJXnhEKaV8m47olVLKx2nQK6WUj9OgV0opH6dBr5RSPs6tJRBEZDzwCtap3W8aY16o8Ph9wGysQ87ygHuNMTvsx54E7rYfe8gYs7S692rbtq2Ji4ur4cdQSin/lpSUdNQY066yxy561I2IBAIpwNVYa31sAKaXBrndpoUx5pR9exLwgDFmvIj0wTo2OQHrDMNlQE9jTElV7zd48GCTmJhYk8+nlFJ+T0SSjDGDK3vMnambBGCPMWafMaYQ+ACYXL5BacjbmmGdHYjd7gNjTIExZj/WKeQJNf0ASimlas+dqZso4EC5+xnA0IqNRGQ28CgQAowt99y1FZ4bVatKlVJK1YrHdsYaY14zxnQDfoW1tKvbROReEUkUkcTs7GxPlaSUUgr3RvQHgc7l7kfb26ryAfB6TZ5rjPk78Hew5ujdqEkp1cCKiorIyMggPz/f6VL8WlhYGNHR0QQHB7v9HHeCfgPQQ0S6YIX0bcCM8g1EpIcxZrd993qg9PYSYL6I/AlrZ2wPYL3b1SmlvEZGRgbNmzcnLi4Oa2031dCMMeTk5JCRkUGXLl3cft5Fg94YUywiPweWYh1e+ZYxZruIzAUSjTFLgJ+LyFVYF2c4Dsy0n7tdRBYAO4BiYHZ1R9wopbxXfn6+hrzDRIQ2bdpQ0ylut46jN8Z8DnxeYdtT5W4/XM1znweer1FVSimvpCHvvNr8N/CdM2PzT8E3v4McvYqbUkqV5ztBX3QW1v4Vlj938bZKKeVHfCfom0fCsAdg+2I4tNHpapRSHnbixAn++te/1vh5EyZM4MSJE9W2eeqpp1i2bFktK6tceHi4R1+vLnwn6AFGPgRNWsM3c52uRCnlYVUFfXFxcbXP+/zzz2nVqlW1bebOnctVV11Vl/K8mls7YxuNsJYw+jH46jew7zvoernTFSnlk579dDs7Dp26eMMa6NOpBU/f0LfKx5944gn27t1LfHw8wcHBhIWFERERwY8//khKSgpTpkzhwIED5Ofn8/DDD3PvvfcCEBcXR2JiInl5eVx33XWMGjWK1atXExUVxSeffEKTJk2YNWsWEydOZOrUqcTFxTFz5kw+/fRTioqK+Oijj+jduzfZ2dnMmDGDQ4cOMXz4cL7++muSkpJo27ZttZ/LGMMvf/lLvvjiC0SEOXPmMG3aNA4fPsy0adM4deoUxcXFvP7664wYMYK7776bxMRERISf/vSnPPLII3X+bn1rRA8w5B5oEQ3LngG9TKJSPuOFF16gW7dubNq0iZdeeonk5GReeeUVUlJSAHjrrbdISkoiMTGRV199lZycnAteY/fu3cyePZvt27fTqlUrFi1aVOl7tW3bluTkZO6//37++Mc/AvDss88yduxYtm/fztSpU0lPT3er7sWLF7Np0yY2b97MsmXLePzxxzl8+DDz58/n2muvPfdYfHw8mzZt4uDBg2zbto2tW7dy11131fLbOp9vjegBgsPgyifhk9mw4xPoO8XpipTyOdWNvBtKQkLCeScNvfrqq3z88ccAHDhwgN27d9OmTZvzntOlSxfi4+MBuOyyy0hNTa30tW+66aZzbRYvXgzAqlWrzr3++PHjiYiIcKvOVatWMX36dAIDA4mMjOTyyy9nw4YNDBkyhJ/+9KcUFRUxZcoU4uPj6dq1K/v27ePBBx/k+uuv55prrnH7+6iO743oAQZOh3a9YfnvoKT6+TulVOPUrFmzc7e//fZbli1bxpo1a9i8eTODBg2qdKmG0NDQc7cDAwOrnN8vbVddm7oaM2YMK1euJCoqilmzZjFv3jwiIiLYvHkzV1xxBW+88Qb33HOPR97LN4M+IBDGPQU5e2Dju05Xo5TygObNm5Obm1vpYydPniQiIoKmTZvy448/snbt2krb1cXIkSNZsGABAF999RXHjx9363mjR4/mww8/pKSkhOzsbFauXElCQgJpaWlERkbys5/9jHvuuYfk5GSOHj2Ky+Xi5ptv5rnnniM5Odkjtfve1E2pXhOg81D47g8wYBqENHW6IqVUHbRp04aRI0fSr18/mjRpQmRk5LnHxo8fzxtvvMEll1xCr169GDZsmMff/+mnn2b69Om8++67DB8+nA4dOtC8efOLPu/GG29kzZo1DBw4EBHhxRdfpEOHDrzzzju89NJLBAcHEx4ezrx58zh48CB33XUXLpcLgN///vceqf2iV5hqaB69wlTaavjXdXDVMzCq7nuulfJnO3fu5JJLLnG6DMcUFBQQGBhIUFAQa9as4f7772fTpk2O1FLZf4vqrjDluyN6gNgR0ONaWPV/cNksaOLezhOllKooPT2dW2+9FZfLRUhICP/4xz+cLsltvh30YM3VvzHKCvur9UQqpVTt9OjRg40bzz/rPicnh3Hjxl3Q9ptvvrngiB8n+X7Qd+gHA26FdX+DhP+ClnolQ6WUZ7Rp08ax6Zua8M2jbiq68tfgKoHvXnC6EqWUanD+EfQRcTDkbtj4b8hOcboapZRqUP4R9ACjfwHBTa2TqJRSyo/4T9CHt4MRD8LOJZCR5HQ1SinVYPwn6AGGz4ambWHZ07rgmVI+rrr14FNTU+nXr18DVuMs/wr60OYw5nFI/R72fuN0NUop1SB8//DKigbfBWtfs5Yx7joWAvyrr1PKI754Ao5s9exrdugP11V9ZNwTTzxB586dmT17NgDPPPMMQUFBrFixguPHj1NUVMRzzz3H5MmTa/S2+fn53H///SQmJhIUFMSf/vQnrrzySrZv385dd91FYWEhLpeLRYsW0alTJ2699VYyMjIoKSnht7/9LdOmTavTx24I/pdyQaFw5Rzrf9Lti52uRinlpmnTpp1bVAxgwYIFzJw5k48//pjk5GRWrFjBY489Rk2XdXnttdcQEbZu3cr777/PzJkzyc/P54033uDhhx9m06ZNJCYmEh0dzZdffkmnTp3YvHkz27ZtY/z48Z7+mPXC/0b0AP1vgdWvWkfgXDIJgkKcrkipxqWakXd9GTRoEFlZWRw6dIjs7GwiIiLo0KEDjzzyCCtXriQgIICDBw+SmZlJhw4d3H7dVatW8eCDDwLQu3dvYmNjSUlJYfjw4Tz//PNkZGRw00030aNHD/r3789jjz3Gr371KyZOnMjo0aPr6+N6lP+N6MGarhn3NBxPheR3nK5GKeWmW265hYULF/Lhhx8ybdo03nvvPbKzs0lKSmLTpk1ERkZWug59bcyYMYMlS5bQpEkTJkyYwPLly+nZsyfJycn079+fOXPmMHdu41hWxT+DHqDH1RA7Er57EQrynK5GKeWGadOm8cEHH7Bw4UJuueUWTp48Sfv27QkODmbFihWkpaXV+DVHjx7Ne++9B0BKSgrp6en06tWLffv20bVrVx566CEmT57Mli1bOHToEE2bNuWOO+7g8ccf99h68fXNf4NexFq++HQWrH3d6WqUUm7o27cvubm5REVF0bFjR26//XYSExPp378/8+bNo3fv3jV+zQceeACXy0X//v2ZNm0ab7/9NqGhoSxYsIB+/foRHx/Ptm3b+MlPfsLWrVtJSEggPj6eZ599ljlz5tTDp/Q8316P3h3vz4D9K+HhzdDMe1abU8rb+Pt69N6kpuvRuzWiF5HxIrJLRPaIyBOVPP6oiOwQkS0i8o2IxJZ7rERENtk/S2r4eerfuKeg6DR8/79OV6KUUvXiokfdiEgg8BpwNZABbBCRJcaYHeWabQQGG2POiMj9wItA6cGlZ40x8Z4t24Pa94aBM2DDP2DYfdAqxumKlFIesnXrVu68887ztoWGhrJu3TqHKnKGO4dXJgB7jDH7AETkA2AycC7ojTEryrVfC9zhySLr3ZVPwtaPYMXv4Uadr1eqKsYYRMTpMtzWv3//RrFefE3UZrrdnambKOBAufsZ9raq3A18Ue5+mIgkishaEZlS4wobQstoSPgZbH4fMndcvL1SfigsLIycnJxaBY3yDGMMOTk5hIWF1eh5Hj1hSkTuAAYDl5fbHGuMOSgiXYHlIrLVGLO3wvPuBe4FiIlxaOpk9GOQPM86iWr6+87UoJQXi46OJiMjg+zsbKdL8WthYWFER0fX6DnuBP1BoHO5+9H2tvOIyFXAb4DLjTEFpduNMQftf/eJyLfAIOC8oDfG/B34O1hH3dToE3hK09Yw8iFY/hykr4WYYY6UoZS3Cg4OpkuXLk6XoWrBnambDUAPEekiIiHAbcB5R8+IyCDgb8AkY0xWue0RIhJq324LjKTc3L7XGfYAhEdaC57pn6dKKR9x0aA3xhQDPweWAjuBBcaY7SIyV0Qm2c1eAsKBjyocRnkJkCgim4EVwAsVjtbxLiHN4PJfQvoaSFnqdDVKKeUResJURSVF8FoCBIXBfasgINC5WpRSyk11PmHKrwQGw9g5kLXDOuRSKaUaOQ36yvS5EToOhOXPQ3HBxdsrpZQX06CvTOkyxifTIfEtp6tRSqk60aCvSrex0GUMrHwJ8k85XY1SStWaBn1VSpcxPpMDa/7idDVKKVVrGvTViboM+kyG1X+BvKyLt1dKKS+kQX8xY5+C4nxrCkcppRohDfqLadsdLr0TEv8Fx/Y7XY1SStWYBr07Ln/COnFqxf84XYlSStWYBr07WnSEofdZJ1Ad2ep0NUopVSMa9O4a9d8Q1gKWPet0JUopVSMa9O5qEgGjHoU9X0PqKqerUUopt2nQ18TQ/4LmneDrp3UZY6VUo6FBXxPBTeCKJ+BgIvz4mdPVKKWUWzToayr+dmjbE775HZQUO12NUkpdlAZ9TQUGwdjfwtFd1sXElVLKy2nQ18YlN1jLI3z7eyg663Q1SilVLQ362ihd8OzUQVj/D6erUUqpamnQ11aXMdBtHHz/v3D2hNPVKKVUlTTo6+KqpyH/BPzwitOVKKVUlTTo66LjQOg3Fda+DrlHnK5GKaUqpUFfV2N/A64i+O4PTleilFKV0qCvq9Zd4bJZkPQO5Ox1uhqllLqABr0njPklBIXC8t85XYlSSl1Ag94TmkfC8Nmw/WM4tNHpapRS6jwa9J4y4iFo0hqWPeN0JUopdR4Nek8JawFjfgH7voW9K5yuRimlznEr6EVkvIjsEpE9IvJEJY8/KiI7RGSLiHwjIrHlHpspIrvtn5meLN7rDL4bWna2RvW6jLFSyktcNOhFJBB4DbgO6ANMF5E+FZptBAYbYwYAC4EX7ee2Bp4GhgIJwNMiEuG58r1McBhc+Ws4vAl2/MfpapRSCnBvRJ8A7DHG7DPGFAIfAJPLNzDGrDDGnLHvrgWi7dvXAl8bY44ZY44DXwPjPVO6lxowDdpdYi9jXOR0NUop5VbQRwEHyt3PsLdV5W7gi1o+t/ELCIRxT8GxvbDxXaerUUopz+6MFZE7gMHASzV83r0ikigiidnZ2Z4syRm9roPOw+DbP0DhmYu3V0qpeuRO0B8EOpe7H21vO4+IXAX8BphkjCmoyXONMX83xgw2xgxu166du7V7r9JljPOOwLrXna5GKeXn3An6DUAPEekiIiHAbcCS8g1EZBDwN6yQzyr30FLgGhGJsHfCXmNv832xw6HneFj1Cpw55nQ1Sik/dtGgN8YUAz/HCuidwAJjzHYRmSsik+xmLwHhwEcisklEltjPPQb8Dquz2ADMtbf5h3FPQcEpWPV/TleilPJjYrzseO/BgwebxMREp8vwnI/vs5ZGeDAZWvr2fmillHNEJMkYM7iyx/TM2Pp2xZNgXNb1ZZVSygEa9PUtItY6Y3bTe5C9y+lqlFJ+SIO+IYz5BQQ3g2/mOl2JUsoPadA3hGZtYcSD8ONncGCD09UopfyMBn1DGT4bmrXTBc+UUg1Og76hhIZbV6JKWwV7vnG6GqWUH9Ggb0iXzYJWsdao3uVyuhqllJ/QoG9IQSEwdg5kboVti5yuRinlJzToG1q/qRDZH1Y8B8WFTlejlPIDGvQNLSAArnoajqdC0ttOV6OU8gMa9E7ofhXEjoKVL0JBntPVKKV8nAa9E0qXMT6dDWv/6nQ1Sikfp0HvlM5DoPdE+OFVOH3U6WqUUj5Mg95J456CotPw/f86XYlSyodp0DupXS+InwEb3oQT6U5Xo5TyURr0TrviSUBgxf84XYlSykdp0DutZTQMvRc2fwCZ252uRinlgzTovcGoRyG0hS5jrJSqFxr03qBpaxj1MKR8CWlrnK5GKeVjNOi9xdD7IDxSlzFWSnmcBr23CGkGl/8KDqy1RvZKKeUhGvTe5NKfQOtusOxZcJU4XY1Sykdo0HuTwGBrGePsnbDlQ6erUUr5CA16b9NnCnSMt46rL8p3uhqllA/QoPc2AQHWgmcnD0DiP52uRinlAzTovVG3K6HrFbDyj5B/yulqlFKNnAa9txr3NJw9Bqv/7HQlSqlGToPeW0Vdas3Xr3kN8rKcrkYp1Yi5FfQiMl5EdonIHhF5opLHx4hIsogUi8jUCo+ViMgm+2eJpwr3C2N/C8X58N2LTleilGrELhr0IhIIvAZcB/QBpotInwrN0oFZwPxKXuKsMSbe/plUx3qrZIxh7qc7SEw9hvGVM0vbdreOrU/6Fxzb53Q1SqlGyp0RfQKwxxizzxhTCHwATC7fwBiTaozZArjqoUa3pB87w8KkA0x9Yw03/GUVC5MyyC/ygZOOLv8VBATD8uedrkQp1Ui5E/RRwIFy9zPsbe4KE5FEEVkrIlMqayAi99ptErOzs2vw0mVi2zRj7a/H8fyN/SgocvGLjzYz8oXl/HHpLo6cbMTHo7foCMPuh20L4fBmp6tRSjVCDbEzNtYYMxiYAbwsIt0qNjDG/N0YM9gYM7hdu3a1fqOmIUHcPjSWrx4Zw3v3DOXS2Ahe+3YPo/6wnJ/PTyYprZFO64x8GMJa6TLGSqlaCXKjzUGgc7n70fY2txhjDtr/7hORb4FBwN4a1FhjIsLI7m0Z2b0t6TlneHdtKh9sOMBnWw7TP6ols0bEMXFgR0KDAuuzDM9p0gpGPwpfPwX7v4cuo52uSCnViLgzot8A9BCRLiISAtwGuHX0jIhEiEiofbstMBLYUdtiayOmTVN+c30f1j45juem9ONsUQmP2dM6f/pqF5mnGsm0TsK90CIKlj2tyxgrpWpE3JnKEJEJwMtAIPCWMeZ5EZkLJBpjlojIEOBjIALIB44YY/qKyAjgb1g7aQOAl40x1Z7XP3jwYJOYmFiXz1QtYww/7Mnh7dX7+ebHLAJFmNC/I7NGxjGocytEpN7eu86S58GSB+HWd6FPvR3ApJRqhEQkyZ4mv/Axb5uzru+gLy8t5zTz1qSxYMMBcguKGRjdklkj45jQ30undUqK4fXh1oj+gbUQ6M7Mm1LKH2jQX8TpgmIWJ2fwr9Wp7Ms+TdvwUG4fGsPtQ2No3yKsQWu5qJ2fwod3wA2vwmUzna5GKeUlNOjd5HIZVu05yturU1n+YxbBgcL1/Tsya2QX4ju3cqSmCxgDb14Fpw7BQ8kQ3MTpipRSXkCDvhb2Hz3NvDWpfJSYQV5BMfGdW3HXyDiu69eRkCCHlwja/z28MxGunmsdeqmU8nsa9HWQV1DMoqQM3lmdyr6jp2nXPJQ7hsYyY2gM7ZqHOlfYv2+GjER4eLN1+KVSyq9p0HuAy2VYuTubt1en8u2ubIIDhRsGdGLmiDgGOjGtc3gL/G00jHrEulCJUsqvadB72L7sPOatSeOjxAOcLixhUEwrZo1wYFpn0T2w8zN4aKO1VIJSym9p0NeT3Pwia1pnTRr7j56mffNQ7hgWy/SEBprWObYf/jIEBt0ON7xS/++nlPJaGvT1zOUyfLc7m7d/SOW7lGxCAgOYOLAjd43oQv/olvX75p8/Dhv+CbPXQdse9fteSimvpUHfgPZm5zFvdSoLkzI4XVjCZbERzBoRx/h+HQgOrIdpnbwseCUeelwNt77j+ddXSjUKGvQOOJVfxMLEDN5Zk0pazhkiW4Rypz2t0ybcw9M6y5+HlS/Cz1ZYlyBUSvkdDXoHuVyGb1Oy+NcPqXy/+yghQQFMGtiJWSPi6BfloWmd/FPwajxE9oOZerVGpfxRdUGvi6XUs4AAYWzvSMb2jmRPVi7vrE5jUXIGC5MyGBIXwawRXbimb2TdpnXCWsDoX8DSJ2Hvcug21nMfQCnV6OmI3gEnzxbxUeIB5q1JI/3YGTq2DDt3tE7rZiG1e9HiAvjzYGgaAT/7FgIcPntXKdWgdOrGS5W4DCt+zOLt1ams2mNN60yJt07C6tupFtM6m96H/9wHU9+Cfjd7vmCllNfSoG8Edmfm8vbqVBYnH+RsUQkJXVpz14g4ru4TSZC70zquEnhjFBTnw+z1EBhcv0UrpbyGBn0jcvJMEQsSD/DOmlQyjp+lU8sw7hwex21DOhPhzrTOri/h/Wlw/Z9gyN31X7BSyito0DdCJS7DNzszeXt1Kqv35hAaFMCU+ChmjYzjko4tqn6iMfDWeDi+31oaIaRZwxWtlHKMBn0jt+tILu+sSWVxcgb5RS6GdmnNXSPjuOqSKqZ10tfCW9fC2N/CmF80fMFKqQanQe8jTpwptKZ1Vqdx8MRZolo14c7hsdw2pDOtmlaY1pl/G6T9YC1j3LS1MwUrpRqMBr2PKXEZlu3M5O0fUlmzL4ew4ABuHBTFzBFx9O5gT+tk7oDXR8Dw2XDt884WrJSqdxr0PuzHI6d4xz5ap6DYxfCubZhlT+sEfvIAbFsEDyZBq85Ol6qUqkca9H7g+OlCPkw8wLtrrGmd6IgmPBAfzPT1NyEDboXJrzldolKqHlUX9Hr6pI+IaBbCfZd347vHr+CNOy4lqlUTfr3iFPOKrsK1cT6pO5OcLlEp5RANeh8TFBjA+H4d+fC/hvP5Q6NJ7XM/Z0wou+b/ktvfXMvXOzIpcXnXX3FKqfqlQe/D+nRqwdO3jSFg1MNcG5hI08xkfjYvkSv+uII3v9/HybNFTpeolGoAOkfvDwry4NVBmDbd+WLwP3l7dRrrU4/RJDiQ2DZNCQ0OJDQogDD7X+snkNDggArb7X+DrdthweW2BQVU/jr27aAAQUSc/iaU8lm6TLG/Cw2Hy3+JfP4LJozaxoT7rmH7oZN8sP4AR07lU1DsoqCohFNni87dLih2UVBcQkGRi4JiF4UlrjqVECCc32EEl3Uo53UYFTqPc8+ptBM6/3XOb3/+62gno/yZWyN6ERkPvAIEAm8aY16o8PgY4GVgAHCbMWZhucdmAnPsu88ZY6q93p2O6OtJcSG8NgRCwuG/vq/xMsYulykL/2KX3QFYt/Mr6RgKikvIL7pw23nty73GxV6nrir+hXFBJ+NmhzEoJoIB0S2141Bep04jehEJBF4DrgYygA0issQYs6Ncs3RgFvCLCs9tDTwNDAYMkGQ/93htPoiqg6AQa0mERXfDtoUw4NYaPT0gQGgSEkiTkMB6KrBqxhgKS1wXdg5FLvLPdQplHUZBUQn55/1lUr7zuLDjKShycex0YZWvU3HndZ+OLZgxNIbJ8Z1oHqYrhCrv587UTQKwxxizD0BEPgAmA+eC3hiTaj9Wceh1LfC1MeaY/fjXwHjg/TpXrmqu703ww8uw/DnoM8UK/0ZAROxRdSCENfz7F9udzOmCYpbuyGT+unTm/Gcb//P5TiYN7MSMoTH0j9JRvvJe7vz9HgUcKHc/w97mDreeKyL3ikiiiCRmZ2e7+dKqxgICYNwzcCINkv7ldDWNRlBgAM1Cg2jfIow7h8Xy+UOj+M/skdwwoBOfbDrEpL/8wMQ/r+K9dWnkFRQ7Xa5SF/CKwyuNMX83xgw2xgxu166d0+X4tu7jIG40fPciFOQ6XU2jJCLEd27FH6YOYN1vxvG7yX0pcRl+8/E2Ep5fxpOLt7Al44TTZSp1jjtBfxAov1BKtL3NHXV5rqoPInDVM3DmKKzRZRHqqkVYMHcOj+OLh0fz8QMjmDigI//ZWDrK/15H+corXPSoGxEJAlKAcVghvQGYYYzZXknbt4HPSo+6sXfGJgGX2k2SgctK5+wro0fdNJAP74C9K+ChTRCuf0V50qn8Ij7ZeJD31qXz45FcmoUEMik+ihkJMfSPrsW1gJVyQ50XNRORCViHTwYCbxljnheRuUCiMWaJiAwBPgYigHzgiDGmr/3cnwK/tl/qeWNMtZPDGvQNJDsF/joMwlpA3xthwDToPNQa8SuPMMaw6cAJ5q9L59Mth8gvctE/qiXTE2KYFN+J8FA9jUV5jq5eqSqXtho2/BN+/H9QfBZaxUL/W6zQb9fT6ep8ysmzRXyy6SDzK4zybx8aQ78oHeWrutOgV9UryLXCfsuHsO9bMC7oGG8Ffr+boXmk0xX6DGMMyekneH99Op/Zo/wB0fYof2AnmukoX9WSBr1yX+4R62IlWxbA4U0gAdD1Civ0e0+0llNQHnHybBH/2WiN8ndl5hIeGsTk+E5MT9BRvqo5DXpVO9m7rMDfugBOpENwU+g1wQr9bldCoJ4V6gmlo/z566xRfkGxi4H2KP8GHeUrN2nQq7oxBg6ss6Z2ti2G/BPQtK01rTPgVoi6THfiesjJM0V8vDGD+evTScnMOzfKnzE0hr6ddJSvqqZBrzynuBD2LLNCf9cXUFIArbtao/z+t0Cbbk5X6BOsUf5x3luXzv/bctga5XduxYyEztwwsBNNQ3SUr86nQa/qR/5J2LHECv3UVYCBqMH2TtyboFlbpyv0CSfPFLF4Ywbz16WzO8sa5U8Z1IkZCbH06dTC6fKUl9CgV/Xv5EFrVcwtCyBzG0ggdL/KmtrpNQFCmjpdYaNnjCEp7bg1l7/1MIX2KP/2hBgmDuyoo3w/p0GvGlbmdnsn7kdw6qC1Bv4lN1ih3+VyCGj4pY59zYkzhSxOPsj89ensycqjeWgQUwZFMWNoDJd01FG+P9KgV85wuSDtB2tqZ8cSKDgJ4R3KduJ2HKg7cevIGEOiPcr/f/Yof1BMK+uInQGdHLl+gHKGBr1yXlE+7F5qjfRTloKrCNr2ggG3QP9bISLW6QobvRNnClmUfJD569LYm32a5mFB3GiP8nt30FG+r9OgV97lzDHY8YkV+umrrW0xw61Rfp8p0LS1o+U1dsYYNqQeZ/66ND7fduTcKH9GQgwTdZTvszTolfc6nmbN5W9ZAEd3QUAw9LjGCv2e4yHYgUtK+ZDjpwtZlGwdl7/PHuXfNCiKGUNj6dWhudPlKQ/SoFfezxg4sqVsJ25eJoS2hD6TrNCPHVXjC5qrMsYY1u8/xvz16Xyx9QiFJS4ujWnFjKGxTBzQkbBgHeU3dhr0qnFxlcD+lVbo71wChXnQIgr6T7WO0Y/s63SFjdqx04UsLjfKbxEWxE2XRjNjaAw9I3WU31hp0KvGq/AMpHxhhf6eZeAqhvZ9rVF+/1ugpbuXL1YVGWNYt/8Y89el8+U2a5R/WWwEMxJiuF5H+Y2OBr3yDaePwvaPrcM1MzYAAnGjrFF+n0kQpmvB1Nax04UsSsrg/fXp7Duqo/zGSINe+Z6cvbB1oRX6x/ZCYCj0Gm+FfverISjE6QobJWMMa/cd4/31ZaP8wbERzBgaw4T+Osr3Zhr0yncZAweTraWUty60LnreJMI6TLP08oi6E7dWcvIKWJScwfvrD7D/6GlaNgnmpkuta9/20FG+19GgV/6hpMi6QtaWD60rZhWdgVYx5S6P2MvpChslYwxr9uUwf106S7cfoajEMCQugukJOsr3Jhr0yv8U5JW7POIK+/KIA8tdHrGD0xU2Sjl5BSy05/JTc87QskkwN18azYyhneneXkf5TtKgV/4tNxO2L7ZC/9BG6/KIXS63Qv+SiRCqAVVTLpdh7b4c5q8vG+UnxLVmxtAYxvfroKN8B2jQK1Xq6G7rUM0tH8KJNAhqAr2vtw7X7DZWL49YC0fLjfLTcs7Qqqk1yp+eEEP39nqN4YaiQa9URcbAgfVW4G9fDGePQ9M29sqa0/TyiLXgctlz+evT+coe5Xdr14yekc3pEdmcnpHh9IxsTlybZoQE6Q5yT9OgV6o6xYWw95uyyyMW50NEFyvwB9yql0eshaN5BSxKyiAx7Ti7M3NJO3aG0qgJChDi2jajZ2Q4Pdo3tzuCcLq0bUZwoHYAtaVBr5S78k/Bzk+t0N+/krLLI94KfW+C8HZOV9go5ReVsCcrjz1ZeaRk5pKSmcfurFzSK3QAXdo2Oxf8VicQTpx2AG7RoFeqNk4dsk/KWgCZW63LI8aNslbV7HmtjvQ94GxhCXuzrdBPycxjt90JHDhe1gEEB1odQI/I5vRob03/9IwMJ7aNdgDladArVVeZO6xVNXd9Adk7rW1tupeFfsxw3ZHrQaUdwLnRf2Yuu7Mu7AC6tg2ne2Q4Pe3Rf4/I5sS1aUqQH3YAdQ56ERkPvAIEAm8aY16o8HgoMA+4DMgBphljUkUkDtgJ7LKbrjXG3Ffde2nQK693PBVSvoKULyH1eygphNAW1lE7PcdDj6uhWVunq/RJZwqL2Zt12uoAsnLZk5lHSlYuB46dPdemtAPoEVk2+u8R2ZzY1r7dAdQp6EUkEEgBrgYygA3AdGPMjnJtHgAGGGPuE5HbgBuNMdPsoP/MGNPP3WI16FWjUpAH+7+zQj/lK8g7AghED7ZG+j2uhQ799QieenamsJg9WXnstoN/d6b110DG8bIOICQwgK7trCmgnu2t8O8RGe4zHUBdg3448Iwx5lr7/pMAxpjfl2uz1G6zRkSCgCNAOyAWDXrlL4yBw5uta+LuXgoHk6ztLaKsq2b1vNY6USukqbN1+pHTBcX2FFDp/L81FXTwRLkOICiArvZO4NLRf4/21j6AwIDG00HXNeinAuONMffY9+8Ehhpjfl6uzTa7TYZ9fy8wFAgHtmP9RXAKmGOM+b6S97gXuBcgJibmsrS0tBp/SKW8Tm4m7PnaGu3vXWFdQCUoDLqMKQv+VjFOV+mXThcUnzsCaHfpv5V0AN3ahduHgYbb5wI0J6Z1U6/sAJwM+lwg3BiTIyKXAf8B+hpjTlX1fjqiVz6puADSVluj/ZQv4fh+a3v7vtDzGmtuP3oIBOjSAU7KK98B2J1AxQ4g1O4ASvcBlB4J1NnhDqC6oA9y4/kHgc7l7kfb2yprk2FP3bQEcozVixQAGGOS7A6gJ6BJrvxLUCh0u9L6Gf97yNljz+svhdV/hlX/B01aQ/errJF+93HWcsuqQYWHBhHfuRXxnVudtz2voNgK/syyQ0E37D/GJ5sOnWsTWv4vAHv03zMynOgI5/8CcGdEH4Q19TIOK9A3ADOMMdvLtZkN9C+3M/YmY8ytItIOOGaMKRGRrsD3drtjVb2fjuiV3zl7AvYuh91fWT9ncqxj9mOGWaHfczy07ak7dL1Qbn4Ru7PyrKN/MnNJybL2BRw+mX+uTVhwaQdg7fzt2d76t3NEUwI82AF44vDKCcDLWIdXvmWMeV5E5gKJxpglIhIGvAsMAo4Btxlj9onIzcBcoAhwAU8bYz6t7r006JVfc5VYO3FTllo/mVut7RFx1hE8Pa+1TtoKCnW0TFW90g6g9ASwlMxc9mTlXdABdG9fGvzWFFCvDtYUUG3oCVNKNVYnM6xRfspS66IqxfkQ3MyaAup5rbVTV9fWbzRO5RdZ0z/lloHYnZnHkVNWB9AvqgWfPTi6Vq+tQa+ULyg6C/u/L5vbP5Vhbe8YX3aGbsd4vXRiI3TybBF7snIpLDYM79amVq+hQa+UrzEGsnaUhf6B9YCBZu3LjuLpeoVeVMWPaNAr5etO58CeZVbw7/kGCk5CQHC5RdiugdZdna5S1SMNeqX8SUkRHFhXNto/mmJtb9uzbFmGmGG6CJuP0aBXyp8d21e2CFvaD/YibC2tY/V7Xgvdr4ZmtZsXVt5Dg14pZSnItY7eSfkSdn8NeZmAQOcEe1mG8RDZV4/Zb4Q06JVSF3K54PCmskXYDm20treILtuh22UMBDdxtEzlHg16pdTF5R4pO2Z/7wooOg1BTayw72mfrNUy2ukqVRU06JVSNVNcAKmrrODf9QWcsFeUjexfNtqPukwXYfMiGvRKqdozxjpyp3RZhvQ1YEqgaRtrR27Pa62razVp5XSlfq2uq1cqpfyZCLTrZf2MfAjOHrcWYSud29/yAQQEWdfNLV2ErU133aHrRXREr5SqPVcJZCSWHbOfZS9qG9HFCvxO8dbyy01bW8suN4mAsFa6TEM90KkbpVTDOHHAGuWnLIX9K61F2C4g1jTPuQ7A7gRKbze1O4TzOojWENJM/0qohk7dKKUaRqvOMOQe66foLJw6ZE31nDkGZ49VfjvvCGTttO4X5lb92oEhFTqFiAodROvKO4igkIb7/F5Kg14pVT+Cm0CbbjV7TnGhFfhVdQpnj9n/nrDO+C3dVlJY9WuGhNsdRKuLdwqlt31sekmDXinlPYJCoHmk9eMuY6DoTBWdwvGyjqN028kM63b+CTCuKl603PRSpX81RFQ+7eSl00sa9Eqpxk3ECtiQZtbUkbtcLmuVzzPlO4cKncK56aVMyP7R6jguOr1U8S+FiEr+aqjQQdTz9JIGvVLKPwUElAVxTZybXqqiUzh3+7j700vBzazQjx4Ct/yrbp+rEhr0SilVE56eXirtFM4eg+Yd66fkenlVpZRSZWo7veQhvrNbWSmlVKU06JVSysdp0CullI/ToFdKKR+nQa+UUj5Og14ppXycBr1SSvk4DXqllPJxXrcevYhkA2l1eIm2wFEPleNJWlfNaF01o3XVjC/WFWuMaVfZA14X9HUlIolVLb7vJK2rZrSumtG6asbf6tKpG6WU8nEa9Eop5eN8Mej/7nQBVdC6akbrqhmtq2b8qi6fm6NXSil1Pl8c0SullCpHg14ppXxcowx6ERkvIrtEZI+IPFHJ46Ei8qH9+DoRifOSumaJSLaIbLJ/7mmgut4SkSwR2VbF4yIir9p1bxGRS72kritE5GS57+upBqqrs4isEJEdIrJdRB6upE2Df2du1tXg35mIhInIehHZbNf1bCVtGvx30s26HPmdtN87UEQ2ishnlTzm2e/LGNOofoBAYC/QFQgBNgN9KrR5AHjDvn0b8KGX1DUL+IsD39kY4FJgWxWPTwC+AAQYBqzzkrquAD5z4PvqCFxq324OpFTy37LBvzM362rw78z+DsLt28HAOmBYhTZO/E66U5cjv5P2ez8KzK/sv5env6/GOKJPAPYYY/YZYwqBD4DJFdpMBt6xby8ExomIeEFdjjDGrASOVdNkMjDPWNYCrUSkfi5eWbO6HGGMOWyMSbZv5wI7gagKzRr8O3OzrgZnfwd59t1g+6fiUR4N/jvpZl2OEJFo4HrgzSqaePT7aoxBHwUcKHc/gwv/Zz/XxhhTDJwE2nhBXQA323/qLxSRhr94ZOXcrd0Jw+0/vb8Qkb4N/eb2n8yDsEaD5Tn6nVVTFzjwndnTEJuALOBrY0yV31cD/k66Uxc48zv5MvBLwFXF4x79vhpj0DdmnwJxxpgBwNeU9diqcslY63cMBP4M/Kch31xEwoFFwH8bY0415HtX5yJ1OfKdGWNKjDHxQDSQICL9GuJ9L8aNuhr8d1JEJgJZxpik+n6vUo0x6A8C5XvdaHtbpW1EJAhoCeQ4XZcxJscYU2DffRO4rJ5rcpc732mDM8acKv3T2xjzORAsIm0b4r1FJBgrTN8zxiyupIkj39nF6nLyO7Pf8wSwAhhf4SEnficvWpdDv5MjgUkikoo1xTtWRP5doY1Hv6/GGPQbgB4i0kVEQrB2VCyp0GYJMNO+PRVYbuy9Gk7WVWEOdxLWHKs3WAL8xD6SZBhw0hhz2OmiRKRD6bykiCRg/f9a7+Fgv+c/gZ3GmD9V0azBvzN36nLiOxORdiLSyr7dBLga+LFCswb/nXSnLid+J40xTxpjoo0xcVg5sdwYc0eFZh79voJq+0SnGGOKReTnwFKsI13eMsZsF5G5QKIxZgnWL8O7IrIHa2ffbV5S10MiMgkotuuaVd91AYjI+1hHY7QVkQzgaawdUxhj3gA+xzqKZA9wBrjLS+qaCtwvIsXAWeC2BuiwwRpx3Qlsted3AX4NxJSrzYnvzJ26nPjOOgLviEggVseywBjzmdO/k27W5cjvZGXq8/vSJRCUUsrHNcapG6WUUjWgQa+UUj5Og14ppXycBr1SSvk4DXqllPJxGvRKeYBYq0ZesAqhUt5Ag14ppXycBr3yKyJyh71G+SYR+Zu96FWeiPyfvWb5NyLSzm4bLyJr7QWvPhaRCHt7dxFZZi8cliwi3eyXD7cXxvpRRN4rd4bqC2KtIb9FRP7o0EdXfkyDXvkNEbkEmAaMtBe6KgFuB5phnZHYF/gO6wxdgHnAr+wFr7aW2/4e8Jq9cNgIoHTpg0HAfwN9sK5LMFJE2gA3An3t13muPj+jUpXRoFf+ZBzWolUb7CUExmEFsgv40G7zb2CUiLQEWhljvrO3vwOMEZHmQJQx5mMAY0y+MeaM3Wa9MSbDGOMCNgFxWMvL5gP/FJGbsJZLUKpBadArfyLAO8aYePunlzHmmUra1XZdkIJyt0uAIHst8QSsi0dMBL6s5WsrVWsa9MqffANMFZH2ACLSWkRisX4PptptZgCrjDEngeMiMtrefifwnX1lpwwRmWK/RqiINK3qDe2141vaSwY/Agysh8+lVLUa3eqVStWWMWaHiMwBvhKRAKAImA2cxrooxRysKxFNs58yE3jDDvJ9lK1QeSfwN3u1wSLglmretjnwiYiEYf1F8aiHP5ZSF6WrVyq/JyJ5xphwp+tQqr7o1I1SSvk4HdErpZSP0xG9Ukr5OA16pZTycRr0Sinl4zTolVLKx2nQK6WUj/v/jYXuVB70eyEAAAAASUVORK5CYII=\n"
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": "<Figure size 432x288 with 1 Axes>",
      "image/png": "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\n"
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plot_loss_curves(history)"
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%%\n"
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "outputs": [],
   "source": [
    "def load_and_prep_image(filename, img_shape=200):\n",
    "  '''\n",
    "  Reads an image from filename, turns it into a tensor and reshapes it to (img_shape, img_shape, color_channels).\n",
    "  '''\n",
    "  #Read in the image\n",
    "  img = tf.io.read_file(filename)\n",
    "  #Decode the read file into a tensor\n",
    "  img = tf.image.decode_image(img)\n",
    "  #Resize the image\n",
    "  img = tf.image.resize(img, size=[img_shape, img_shape])\n",
    "  #Rescale the image (get all values between 0&1)\n",
    "  img = img/255.\n",
    "  return img\n",
    "\n",
    "def pred_and_plot(model, filename, class_names):\n",
    "  '''\n",
    "  Imports an image located at filename, makes a prediction with model and plot th eimage with the predicted class as the title.\n",
    "  '''\n",
    "  #Import the target image and preprocess it\n",
    "  img = load_and_prep_image(filename)\n",
    "\n",
    "  #Make a prediction\n",
    "  pred = model.predict(tf.expand_dims(img, axis=0))\n",
    "  #Get the predicted class\n",
    "  pred_class = class_names[int(tf.keras.backend.eval(tf.math.argmax(input = tf.squeeze(pred))))]\n",
    "\n",
    "  #Plot the image and predicted class\n",
    "  plt.imshow(img)\n",
    "  plt.title(f'Prediction: {pred_class}')\n",
    "  plt.axis(False)"
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%%\n"
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "outputs": [
    {
     "data": {
      "text/plain": "<Figure size 432x288 with 1 Axes>",
      "image/png": "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\n"
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "pred_and_plot(model, 'test_1.png', class_names=class_names)"
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%%\n"
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "outputs": [
    {
     "data": {
      "text/plain": "<Figure size 432x288 with 1 Axes>",
      "image/png": "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\n"
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "pred_and_plot(model, 'test_2.png', class_names=class_names)"
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%%\n"
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "outputs": [
    {
     "data": {
      "text/plain": "<Figure size 432x288 with 1 Axes>",
      "image/png": "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\n"
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "pred_and_plot(model, 'test_3.png', class_names=class_names)"
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%%\n"
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "outputs": [
    {
     "data": {
      "text/plain": "<Figure size 432x288 with 1 Axes>",
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAOcAAAD3CAYAAADmIkO7AAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjUuMSwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy/YYfK9AAAACXBIWXMAAAsTAAALEwEAmpwYAAAfI0lEQVR4nO3de3RU9d3v8fdv7rlnck8gFwnhJgEkIgpKSL1AUEDQerhY2prjEqtoV4+u1vY8fY4+1rZe+ugCD1arKB4pXirSoqDAUaCK4aBFkGANhksgCQkht8nkMpd9/shknkhBbiF7z8z3tRZrTWbPzP7OMJ/5/fbev/3bStM0hBDGY9K7ACHEqUk4hTAoCacQBiXhFMKgJJxCGJSEUwiDknBeJEqpl5VSjwZuX6OU+ud5vs5zSql/69/qRCiI6HAqpQ4qpTqUUi6l1LFAoGL7ez2apm3TNG34WdTzI6XU30967mJN0/6jv2sSxhfR4QyYqWlaLDAeuBz4nyc/QCllGfCqIpB8zt8m4QzQNO0osB4YDaCU0pRS9yilKoHKwH03KaV2KaWalVKfKKXG9D5fKXWZUupzpVSbUup1wNFn2VSl1JE+f2crpd5WSjUopRqVUsuUUiOB54CrAi15c+Cxwe5x4O87lVL7lVInlFJ/VUpl9VmmKaUWK6UqAzU+q5RSZ/P+lVJXKKV2KqVaA72IP/RZ9gOl1KFArb8K9DiuO019J7/XXyilvgl8LhVKqTl9lv1IKfWxUuo/lVKNwP9SStmVUk8qpQ4H6nhOKRV1Nu8h3Eg4A5RS2cAM4B997r4ZmAiMUkpdBrwE3AUkA38E/hr4MtmAd4BXgSTgTeCW06zHDKwDDgF5wCBgtaZp+4DFwHZN02I1TUs8xXO/B/wWuA3IDLzG6pMedhMwARgTeNy0wHNzAoHNOc1H8AzwjKZp8UA+8EbgeaOA5cAPgKzAex98mtc4lW+Aa4AE4GHg/yilMvssnwhUAenAb4DfAcOAccBQej6fX5/D+sKHpmkR+w84CLiAZnq+6P8biAos04Dv9XnscuA/Tnr+P4FiYApQA6g+yz4BHg3cngocCdy+CmgALKeo50fA30+67+U+r/Mi8HifZbGAB8jrU/PVfZa/AfziLD+LrfSEJ+Wk+39Nz49H798xQDdw3cn1nfxeT7OeXcDsPu/3cJ9lCmgH8vvcdxVwQO/vih7/pOWEmzVNS9Q0LVfTtJ9omtbRZ1l1n9u5wP8ItD7NgW5nNj2tSRZwVAt8mwIOnWZ92cAhTdO851FrVt/X1TTNBTTS07r0qutz201PgM9GGT0t1ldKqf+nlLqpzzqDn4Omae2BdZ4VpdSiPpsCzfRsNqT0eUjfzzgViAY+6/P4DYH7I45sgH+3vmGrBn6jadpvTn6QUqoYGKSUUn0CmkNPl+5k1UCOUspyioCe6RShGnp+JHrXG0NPN/PoGZ53RpqmVQLzlVImYC7wllIqGagFRvZZZ3Rgnb3a6QlUr4w+j80FXgCupae77lNK7aKnhQyuus/t40AHcKnWsw8goknLefZeABYrpSaqHjFKqRuVUnHAdsAL3KeUsiql5gJXnOZ1dtDzhf9d4DUcSqnJgWXHgMGBbdhT+TPwY6XUOKWUHXgMKNc07eCFvjml1O1KqVRN0/z0dPMB/MBbwE1KqasDdT3Ct783u4AZSqkkpVQG8NM+y2LoCV9DYB0/JrDD7VQC634B+E+lVFrgOYOUUtMu9P2FIgnnWdI0bSdwJ7AMaAL207PNhKZp3fS0Nj8CTgD/DXj7NK/jA2bSs7PjMHAk8HiA/wvsBeqUUsdP8dxNwL8Bf6En4PnAvLOpP7BDyPUdO4SmA3uVUi56dg7N0zStQ9O0vcA9wKrAOpsCNfd6FfiCnu33D4DX+9RbATxFz4/XMaAQ+PgMpf6cns/2U6VUK7AJOOMx4nCkvr2ZJMSZKaUOAv898GMhLhJpOYUwKAmnEAYl3VohDEpaTiEM6kzHOaVZFeLiO+X4Z2k5hTAoCacQBiXhFMKgJJxCGJSEUwiDknAKYVASTiEMSsIphEFJOIUwKAmnEAYl4RTCoCScQhiUhFMIg5JwCmFQEk4hDErCKYRBSTiFMCgJpxAGJeEUwqAknEIYlIRTCIOScAphUBJOIQxKwimEQUk4hTAoCacQBiXhFMKgJJxCGJSEUwiDknAKYVASTiEMSsIphEFJOIUwKAmnEAYl4RTCoCScQhiUhFMIg5JwCmFQEk4hDErCKYRBSTiFMCgJpxAGJeEUwqAknEIYlIRTCIOScAphUBJOIQxKwimEQUk4hTAoCacQBiXhFMKgJJxCGJSEUwiDknAKYVASTiEMSsIphEFJOIUwKAmnEAYl4TxHmqahaRoejwe/34+maXqXJMKUhPMceTwePv/8c8aNG8fKlSuDYRWiv0k4z9Hx48d56qmnqKiooKysjAkTJvDmm2/idrv1Lk2EGXWGX31pEvpwu918/PHHzJo1C4/HQ2pqKg0NDdhsNkaPHs1PfvITSktLSU9P17tUEVrUKe+UcJ69ffv28cgjj/Dmm2+Sl5fHwYMHcTqdmEwmWltbiY2NZezYscyZM4fS0lKGDBmid8kiNEg4L0RLSwvr1q1jyZIluFwuLBYLHR0dWK1WlFLEx8fj8XjQNI3ExERGjBjBddddx6233soll1yid/nC2E4ZTstAVxGq9u/fz/r162lrayM1NZXa2lqgZwcRQHt7O36/H4fDQUtLCzt27KCqqoodO3ZQUlLCjTfeSG5urp5vQYQYCedZaGtr4x//+AdbtmxBKUVnZ+e/PKajowMAk8mEyWRCKUVDQwMbNmxg37597Nq1i5KSEq6//nqSk5NR6pQ/lkIESTjPwr59+9iyZQvHjh0jOjqapqam0z62N6QOhwObzUZXVxeHDx+msrKSzz//nIMHD3LllVdSWFiI0+nEbDYP1NsQIUa2Oc/A5XKxfPlyli5dSmNjI3FxcRw7duysn2+xWEhJSaGlpQWlFN3d3RQXF3PLLbdw+eWXk5+fT0JCgoQ0sskOofNRXl7O73//e959911sNhsul+u8XysvL4+jR4+ilMLv91NSUsKCBQuYOHEimZmZxMXFSUgjk4TzXHk8Hh599FFeeuklWlpaiImJoa6u7oJfNz8/n+rqanw+H2azmQkTJrBgwQJKS0vJyMjAbrdjMsn4kAgi4TxX5eXlPPTQQ3zyySen3RF0IXJzc2lqaqKtrQ2LxcKIESMoKytj4cKFJCYmBltR2XkU9iSc50LTNO644w7Wrl0LgNls5vjx4xdlXampqSilqK+vx2w2k56ezuLFi7n33ntJTEwMPk5CGrYknOdi48aN3H///Rw+fBifz9fvreapxMTEkJKSwqFDhzCZTGRkZPDggw9SVlZGbGyshDN8STjPxdy5c9m4cSN2ux2Px0Nra+uArFcphclkIjs7m0OHDmGz2Rg0aBB33XUXixYtIi0tTbZHw4+E82ytW7eOu+++m/b2djo7O+ns7Bzw08IsFgter5ecnBxqamqIjY1l2LBh3HLLLcyePZuCggIJafiQcJ6Jpmn4fD5uvvlmNm/eTHR0NG63e0C6tKdjt9uxWCzExcXR1tZGQkICQ4cOZerUqcyePZvx48frVpvoNxLOM9E0jY8++oh58+bR3d1Nd3c3nZ2d+P1+XeuyWCzYbDYAbDYbmqaRkJDA6NGjufrqqyktLWXcuHG61iguiITzu2iaht/vp6ysjFWrVpGYmEhzc3NwYLsRWK1WbDZbsDtrNptxOp1cccUVTJ48menTpzN06FDZcRR6JJzfRdM0du3aRWlpKW63G6UU7e3t+Hw+vUv7F3a7HbvdjqZpeL1eANLS0pg9ezbf+973KCoqIjMzU0YbhQ4J5+n0Ttj1wAMPsGzZMjIzMzl27Jghg9mX1WolJiYGr9eL1+vF5/MxYsQI5s+fzzXXXMOwYcNwOp1YrVa9SxXfTcJ5On6/n+rqaoqKimhtbcVms+F2u0Nm4i6TyUR6enpw77LX66WwsJA777yTq6++muzsbOLi4iSkxiXhPBVN03C73Tz11FP8+7//O1lZWdTW1oZMME+Wm5tLY2MjnZ2d+Hw+xo8fz/z585k+fTp5eXk4HA7p7hqPhPNUfD4fBw8eZNKkSTQ0NIRsKE+Wm5tLS0sLLpcLr9fLsGHDmDdvHosWLSInJwez2SzHSY1DwnkyTdNobGzk8ccf54knnmDIkCFUVVXpXVa/GjRoEJ2dnTQ3N+P3+0lLS+O2227j/vvv/9YEZLKHV1cSzpN5PB52797NtddeS0tLi97lXFROp5Po6Ohglz0qKoqysjIeeughMjIyJJz6OuWHH9H9mqNHj/Lqq6/icrnIy8vTu5yLqqmpiZqaGmw2G0OGDKGrq4vly5czevRoHnzwQaqqqgx1TFdEcMvZ0dHB5s2bueOOO2hoaAiOZY0EFosFn88X7MbHxMSQnp7OvHnzmD9/PgUFBcERSWJASLe2rz179rBs2TJefvll0tLSOHLkiN4lDTiHw0FnZyfZ2dnU1dXhdDrJycmhpKSEWbNmUVRURFRUlN5lRgKZt7ZXR0cHu3fv5v3338fj8XznbHrhrHdAf1NTU3AQQ0VFBTU1NezcuZOrrrqK0tJSJk6cKMdIdRCR4fz666/ZunUrx44dw+l0cuLECb1L0lXvpGVdXV10d3fT0tLC559/TmVlJXv27OHKK6+kpKSEK664Ijgnr7j4Ii6cbrebTz75hK1bt+L3++WL1kdvSHtPMK+rq+PDDz+kvLycPXv2UFpayoQJExg2bJiEdABE3Dbn7t27efzxx3nrrbcwm81y6b7vYLfbiY6OxuVyYbVaSUxM5Pvf/z4zZsxgxIgRZGZmSne3f8gOoe7ubpYuXcoLL7xATU1Nv011Gc5MJhMxMTHExsbS0NAAQEZGBmVlZcHrvyQmJsre3Qsj4aysrOSBBx5gw4YNwTNRxNlRSpGSkoLJZAqGdMiQISxcuJCZM2dSUFBAdHQ0FkvEbSn1h8gOp9/v5+GHH+aVV14JXsLvXC6rIP5LWloafr+f5uZmvF4vgwcP5tZbb2XRokVceumlWCwWGbd7biI7nBUVFSxatIhdu3YBGP5czVCQmZmJ1+ultbWVrq4u0tLSmDZtGnfffTdXXnklIGN2z1Jkh/Oee+7hjTfewGKx4PF4aGxs1LuksJGamorZbKa5uZnOzk6io6O59tpreeSRRxg7dqwE9MwiN5xffPEFN998M/X19Xg8HtnWvEhiY2ODx407Ozux2WzMmjWLn//85xQWFsr26OlFbjgXLlzI2rVriY+Pp62t7YKuFCZOr3dCbJPJxODBgzly5Ahms5m4uDjmzJnDXXfdxciRI2VI4L+KzHB+9tlnzJw5k5aWFnw+Hx6PR/epLsOdUgqbzUZ3dzc5OTnU1tYSExPDyJEjeeaZZxg/frzsMPq2yDxl7MUXX+TEiRPEx8cDSDAHgKZpdHV1BU9mT0pKorm5mfz8fOLi4mQb9CyF7UaApml89dVXvPfee1gsFtxud8ScEmYkXq8Xt9tNUlISt9xyC1lZWRLOsxS2Laff72fVqlVUV1cTExNj2Dlow118fDytra0UFxdTVFREbGys3iWFjLAMp9/vp7a2ltdeew2LxSJdWZ30DumLiYnh9ttvx+l0Sqt5DsIynB6Ph3feeYcDBw4QGxtLY2OjtJoDTClFfHw89fX1XHbZZRQXFxMdHa13WSEl7MLp9/tpampi+fLlmEwmmaNVJ72HVKxWKz/96U/l4r/nIazCqWkabW1tvP7661RUVJCenh5Wc9GGCrPZTHJyMvX19YwbN45Zs2bJWSvnIezC2dDQwDPPPAPIYRM99bacv/zlL6X3cp7CKpz19fX85S9/4dChQ+Tk5MhZJzqw2+2kp6cHW83Zs2dLd/Y8hU04fT4fVVVVPP/882iaFvaTRBuVz+ejvb0dh8PB7373O0DOTDlfYRPOqqoq1qxZw9GjR8nKyqK1tVXvkiJOVFQUTqeT7u5uJk2axJQpU/QuKaSFxQghr9fL3r17WbNmDR6PB5fLJTuBdODz+ejs7CQuLo7Fixdjt9v1LimkhUXLWVlZydatW6mrqyM5OVnOOtFBbGxs8HDJ+PHjKSkp0bukkBfyLWd3dzfl5eVs2rSJzs5OlFIy4EAHfr8fr9dLSkoKCxYsICkpSe+SQl7It5z79+/n448/5sCBAzgcDmk1dRAdHR0c8DFmzBhuuOEGvUsKCyEdTq/Xy9atW9mxYwder1e2cXRis9kwmUxkZmYybdo00tPT9S4pLIR0OGtqati6dSv79+/H5/NF/GUV9GC1WnG73fh8PsaMGcOMGTP0LilshGw4fT4f69at44svvsBut+N0OvUuKSKlpKTgcDgYPnw4N9xwAzk5OXqXFDZCdodQfX09b775Jl9//bWcRK0TpRQNDQ2YzWYmT54srWY/C8mWU9M0VqxYwf79+0lJSSElJUXvkiLSJZdcgt1up6ioiOLiYtnW7Gch2XLW19fz2muvUVdXh8/nkwEHOjl48CBKKW688UaKi4v1LifshGTL+Yc//IGamhoyMzODE3eJgTV06FAsFgs33XQTxcXFJCcn611S2Am5lvPAgQOsWrUKl8tFW1ubnBamk0OHDmEymbj11lspLCyUwe0XQci1nM8//zyNjY2kp6djtVqlS6uD3Nxc/H4/06ZNo7CwUCbtukhCpuXUNI3a2lreeOMNfD4fTU1NdHd3611WRGpoaMDhcDBv3jxyc3NlguiLJGQ+VU3TWLVqFYcPH8bpdNLV1SVdWh2kp6fjdruZOnUql19+OXFxcXqXFLZCIpx+v59jx46xcuVKTCaTHNfUkc/nw263s2DBAjIyMmQKkosoJMLp9XrZuHEje/bsIT4+nhMnTsiZJzqIiYnh+PHjFBYWUlxcTExMjN4lhTXDh7N3qss//elPmEwm7Ha77BnUgVKKxMRErFYrP/7xj0lKSpJtzYvM8FcZc7vdvP/++8ydO5eEhASZG0gnZrMZn8/HsGHDKC8vJyEhQX4k+0/oXWWs9ypVTzzxBEopOdCtE4vFQlZWFiaTifvuu0+uFDZADN1yNjc38+677/KDH/yAxMREmpqa9CwnIpnNZqKjo3G5XGRlZVFZWYnD4ZBw9q/Qazmrq6tZunQpgOyy14nFYiEhIQGbzcYvfvELCeYAMmw4a2tr2bRpEzt37iQhIYHDhw/rXVLEsVqt2Gw26urqGDZsGHfeeafeJUUUw44Q+vLLL1mxYgWATD+iE7PZjM1mw2q1smTJErneyQAzZMt58OBBtm3bRlVVFYmJiTQ2NupdUsSJiorCZrPR0dFBQUEBc+fORSklXdoBZMhwlpeX89e//pWOjg48Hg9paWl6lxRxei9ElJyczMKFC2UaGB0YMpy5ublcffXV5Ofn43a7g9feSE1N1bu0iBAdHR0cljdq1ChmzZolAw50YMhtztGjR1NWVsbIkSP5+OOP2b59OzU1Nfh8Pmw2G1FRUXIu50VksVjQNI2UlBSuv/56cnNz9S4pIhkynLGxsVx22WWMHDmSqVOnsmbNGj788EO+/PJLTCYTNpsNm82GUgqPxyMD4fuR3W7H5/NhNpsZPXq0TNqlI0OGs5fD4eDSSy9lxIgRlJaW8vLLL7NlyxYOHTqE1WrF4XDgdruBntFEMhj+wsXGxtLZ2UlmZibXXHMNI0eO1LukiGXocPYym80UFRVRVFREeXk5Tz/9NJ9++in19fWYTCbi4+Pxer20t7ejaZp0d8+T2Wymq6sLgEmTJjFz5kydK4psIbeVP3HiRP785z+zYsUK5syZg9PpxOVy4fV6SU1Nxel0yu7+85SZmYnP5yM3N5fJkydTUFCgd0kRzdBja8/Gzp07WbFiBWvWrKGurg6z2Ux6ejptbW1yAd1zFB0dTXd3N0uWLOHee+9lyJAhepcUKU7ZmoR8OHu7sd988w2rV6/mxRdfDA71y8vLo6mpidbWVpkI7Azy8vKoqamhsLCQhx9+mBkzZkgPZOCEZziBYPB8Ph+NjY28+uqrPPnkkxw7dgzoOW5aU1ODx+PRs0xDczgcdHV18fTTTzN//nw5pjywQu+slLPVO6zMbDaTmprKPffcw2effcajjz5KRkYGR44cwePxkJGRIVNrnEJ+fj4ej4fJkyczceJEOW/WIMKi5TyZpmlomkZzczO1tbW89NJLvP766xw/fpzu7m4SEhKwWCwcP35c71INITo6mo6ODp577jnmzp0r154ZeOHbrf0uPp+P6upq9u/fz4YNG1i/fj0HDx7E6/WilMLpdFJXV6d3mbrJzs7m6NGjTJkyhWXLljF8+HAslpA4whZOThnOsP9fMJvN5OXlkZWVRU5ODhMnTmTbtm189NFHfP3117hcrmB3OBJD2tbWBsDtt9/O4MGDJZgGEjH/EzabjWHDhpGTk8OoUaO47LLL2L59Ozt27GDv3r3BljQpKYkTJ05ExN7dlJQUGhsbGTt2rEx1aUARE85evUMCCwoKmDJlChs3bmTDhg3s2bOHpqYmbDYbJpOJqKgoOjo6wnZIoFIqOD55/vz5ZGZmSqtpMGG/zXkmHo+H6upqVq9ezd/+9jeqqqpoamoiNTWVpqYmfD4ffr8/rAbXK6WCh05yc3PZvHkzOTk5Mnu7fsL3UMqFsFqtDBkyhIceeojVq1ezaNEisrOzaWtrw2Qy4XQ6iYmJwWw2h82X12QyBU9g/+EPf0h2dnbYvLdwEvEt56ns3buX5cuXs27dOurq6lBKERsbi8ViCQ5sCOVtUpPJhN/vJzExka+++oq0tDQZDaSvyDyUciEqKyt58cUXeeuttzhw4ABms5n4+HiioqI4cuSI3uWdF7vdzqBBg6iurua+++4LTtgtdCXhPFe9n82BAwd45513eOWVV9i9e3fw+Gh0dHRIhlQpRUZGBrt37yY5OVnCqT8J5/nqHXHU2NjIBx98wG9/+1v27t0L9Jxm5ff7aWhoMPx5pDExMSQmJtLW1saDDz7Ir371KwmmMUg4L1TvXtvW1lbeffddnnzySSoqKoLz7Xg8nuC5pUZlsVjIz89ny5YtpKen612O6CHh7C+apuFyuThx4gRr1qzh2Wefpbq6Go/HE5xC0uVyBWcVMIK4uDhsNhtms5nFixfz61//WvbQGoeEs79pmkZDQwNVVVV88MEHvP322xw4cICOjg7sdjs2m40TJ07oXWYwlCaTiXHjxrFy5Uo5kdpYInNs7cWklCItLY2UlBQyMzOZMGECW7ZsYdOmTVRWVuJ2uzGbzTidTl3PgImKikLTNJxOJ9OmTZOpLkOEhLMfmEwmcnNzyc7OZujQoYwdO5ZPP/2U8vJyKioqgid5JycnD/ilJaxWKz6fD5PJxPDhw5k5c6Z0Z0OEhLMfmUwmCgoKGDJkCJMmTWLbtm1s3ryZnTt30tHRgdVqBSA+Pp62trYBGcjgcDiCJ5pPnjyZ0aNHX/R1iv4h4bwIzGYzubm5DB48mOuuu461a9fyzjvv8NVXX2GxWIiLi8PtdmOxWPB4PBdtcL3VasVkMmG1Wrn88suZPn26DG4PIRE/tvZiMpvNZGRkcOedd7Jq1Sp+9rOfMWbMGDo6OlBKkZKSgsPhuGjjduPj4+ns7CQpKYnJkydTVFTU7+sQF4+EcwD0DqC/9957ee+991iyZAk5OTk0NTXR1dVFUlISSUlJKKX67YJBvYMLlFKUlJQwZcoU2dYMMXIoRScNDQ289NJLvPLKK1RWVqJpGrGxsSQmJnLo0KELfv309HQaGxtJT0/nscceY+HChRJO45LjnEZUW1vL2rVrWblyJeXl5Wiaht1uJysri6qqqvN+3b7D9MrKyhg6dGg/Vi36mYTTiPrOFPj3v/+dZ599lg8++ADouahQeno6VVVV57RnNzs7m5qaGvLz83n++eeZMmWKjKE1NjnZ2oh6tzOdTielpaW89tprrF+/nunTp+NyuThw4AAJCQkMGjQoeCjmTNrb2/H7/dxxxx0MHTpUghmipOU0GL/fT1dXF01NTXz22Wc89thj7Ny5E03TiI6OJioqivb2dtrb20/5/NzcXI4cOUJBQQGrVq2isLBQDp8YnwzfCwW9k4s5HA5KSkrIy8tj27ZtrFq1it27d9Pc3IzD4SA5OZmuri5cLte3nt/W1obP5+Ouu+4iJydHghnCpOUMAfX19VRUVPDJJ5/w3nvvsW/fvuC43d6BDG63O7iHNj8/n7fffpuCgoKz7goLXUnLGarS0tJIS0ujoKCASy+9lO3bt7N9+/ZgSHsHtSul8Hq93HbbbWRnZ0swQ5y0nCHoyJEjlJeX89FHH1FeXs4///lPrFYrbreblJQU1q1bx6hRo6RLGzqk5QwXgwcPZvDgwUyZMoUPP/yQNWvW8MUXX/DNN98wZ84cRowYIcEMA9JyhjhN02hvb2f9+vUsXbqUP/7xjwwfPrzfhgGKASGDEMKZpmn4fD7MZrMc1ww9Ek4hDEpGCAkRSiScQhiUhFMIg5JwCmFQEk4hDErCKYRBSTiFMCgJpxAGJeEUwqAknEIYlIRTCIOScAphUBJOIQxKwimEQUk4hTAoCacQBiXhFMKgJJxCGJSEUwiDknAKYVASTiEMSsIphEFJOIUwKAmnEAYl4RTCoCScQhiUhFMIg5JwCmFQEk4hDErCKYRBSTiFMCgJpxAGJeEUwqAknEIYlIRTCIOScAphUBJOIQzKcoblakCqEEL8C2k5hTAoCacQBiXhFMKgJJxCGJSEUwiDknAKYVD/H9mAXbmnwxzeAAAAAElFTkSuQmCC\n"
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "pred_and_plot(model, 'test_4.png', class_names=class_names)"
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%%\n"
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "outputs": [
    {
     "data": {
      "text/plain": "<Figure size 432x288 with 1 Axes>",
      "image/png": "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\n"
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "pred_and_plot(model, 'test_5.png', class_names=class_names)"
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%%\n"
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "outputs": [
    {
     "data": {
      "text/plain": "<Figure size 432x288 with 1 Axes>",
      "image/png": "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\n"
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "pred_and_plot(model, 'test_6.png', class_names=class_names)"
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%%\n"
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "outputs": [
    {
     "data": {
      "text/plain": "<Figure size 432x288 with 1 Axes>",
      "image/png": "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\n"
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "pred_and_plot(model, 'test_7.png', class_names=class_names)"
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%%\n"
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "outputs": [
    {
     "data": {
      "text/plain": "<Figure size 432x288 with 1 Axes>",
      "image/png": "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\n"
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "pred_and_plot(model, 'test_8.png', class_names=class_names)"
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%%\n"
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "outputs": [],
   "source": [
    "model.save('AlexNet.h5')"
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%%\n"
    }
   }
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 2
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython2",
   "version": "2.7.6"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 0
}