densenet121cottondisease.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "color-methodology",
   "metadata": {
    "_cell_guid": "b1076dfc-b9ad-4769-8c92-a6c4dae69d19",
    "_uuid": "8f2839f25d086af736a60e9eeb907d3b93b6e0e5",
    "execution": {
     "iopub.execute_input": "2021-06-08T10:27:33.776643Z",
     "iopub.status.busy": "2021-06-08T10:27:33.775983Z",
     "iopub.status.idle": "2021-06-08T10:27:39.441537Z",
     "shell.execute_reply": "2021-06-08T10:27:39.440935Z",
     "shell.execute_reply.started": "2021-06-07T07:19:59.967664Z"
    },
    "papermill": {
     "duration": 5.686857,
     "end_time": "2021-06-08T10:27:39.441683",
     "exception": false,
     "start_time": "2021-06-08T10:27:33.754826",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [],
   "source": [
    "import tensorflow \n",
    "\n",
    "import pandas as pd\n",
    "import numpy as np\n",
    "import os\n",
    "import keras\n",
    "import random\n",
    "import cv2\n",
    "import math\n",
    "import seaborn as sns\n",
    "\n",
    "from sklearn.metrics import confusion_matrix\n",
    "from sklearn.preprocessing import LabelBinarizer\n",
    "from sklearn.model_selection import train_test_split\n",
    "\n",
    "import matplotlib.pyplot as plt\n",
    "\n",
    "from tensorflow.keras.layers import Dense,GlobalAveragePooling2D,Convolution2D,BatchNormalization\n",
    "from tensorflow.keras.layers import Flatten,MaxPooling2D,Dropout\n",
    "\n",
    "from tensorflow.keras.applications import DenseNet121\n",
    "from tensorflow.keras.applications.densenet import preprocess_input\n",
    "\n",
    "from tensorflow.keras.preprocessing import image\n",
    "from tensorflow.keras.preprocessing.image import ImageDataGenerator,img_to_array\n",
    "\n",
    "from tensorflow.keras.models import Model\n",
    "\n",
    "from tensorflow.keras.optimizers import Adam\n",
    "\n",
    "from tensorflow.keras.callbacks import ModelCheckpoint, ReduceLROnPlateau\n",
    "\n",
    "import warnings\n",
    "warnings.filterwarnings(\"ignore\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "exotic-civilization",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2021-06-08T10:27:39.463072Z",
     "iopub.status.busy": "2021-06-08T10:27:39.462490Z",
     "iopub.status.idle": "2021-06-08T10:27:39.467607Z",
     "shell.execute_reply": "2021-06-08T10:27:39.467203Z",
     "shell.execute_reply.started": "2021-06-07T07:20:01.4383Z"
    },
    "papermill": {
     "duration": 0.016712,
     "end_time": "2021-06-08T10:27:39.467713",
     "exception": false,
     "start_time": "2021-06-08T10:27:39.451001",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Tensorflow-version: 2.4.1\n"
     ]
    }
   ],
   "source": [
    "print(\"Tensorflow-version:\", tensorflow.__version__)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "differential-dryer",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2021-06-08T10:27:39.491620Z",
     "iopub.status.busy": "2021-06-08T10:27:39.491123Z",
     "iopub.status.idle": "2021-06-08T10:27:44.541097Z",
     "shell.execute_reply": "2021-06-08T10:27:44.540536Z",
     "shell.execute_reply.started": "2021-06-07T07:20:01.452185Z"
    },
    "papermill": {
     "duration": 5.065603,
     "end_time": "2021-06-08T10:27:44.541225",
     "exception": false,
     "start_time": "2021-06-08T10:27:39.475622",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Downloading data from https://storage.googleapis.com/tensorflow/keras-applications/densenet/densenet121_weights_tf_dim_ordering_tf_kernels_notop.h5\n",
      "29089792/29084464 [==============================] - 0s 0us/step\n"
     ]
    }
   ],
   "source": [
    "model_d=DenseNet121(weights='imagenet',include_top=False, input_shape=(256, 256, 3)) \n",
    "\n",
    "x=model_d.output\n",
    "\n",
    "x= GlobalAveragePooling2D()(x)\n",
    "x= BatchNormalization()(x)\n",
    "x= Dropout(0.5)(x)\n",
    "x= Dense(1024,activation='relu')(x) \n",
    "x= Dense(512,activation='relu')(x) \n",
    "x= BatchNormalization()(x)\n",
    "x= Dropout(0.5)(x)\n",
    "\n",
    "preds=Dense(4,activation='softmax')(x) #FC-layer"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "golden-throat",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2021-06-08T10:27:44.567940Z",
     "iopub.status.busy": "2021-06-08T10:27:44.567257Z",
     "iopub.status.idle": "2021-06-08T10:27:44.591470Z",
     "shell.execute_reply": "2021-06-08T10:27:44.591037Z",
     "shell.execute_reply.started": "2021-06-07T07:20:06.420879Z"
    },
    "papermill": {
     "duration": 0.039488,
     "end_time": "2021-06-08T10:27:44.591581",
     "exception": false,
     "start_time": "2021-06-08T10:27:44.552093",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [],
   "source": [
    "model=Model(inputs=model_d.input,outputs=preds)\n",
    "#model.summary()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "patent-bradford",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2021-06-08T10:27:44.628557Z",
     "iopub.status.busy": "2021-06-08T10:27:44.627315Z",
     "iopub.status.idle": "2021-06-08T10:27:44.629610Z",
     "shell.execute_reply": "2021-06-08T10:27:44.630061Z",
     "shell.execute_reply.started": "2021-06-07T07:20:06.455076Z"
    },
    "papermill": {
     "duration": 0.028773,
     "end_time": "2021-06-08T10:27:44.630203",
     "exception": false,
     "start_time": "2021-06-08T10:27:44.601430",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [],
   "source": [
    "for layer in model.layers[:-8]:\n",
    "    layer.trainable=False\n",
    "    \n",
    "for layer in model.layers[-8:]:\n",
    "    layer.trainable=True"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "cubic-survivor",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2021-06-08T10:27:44.656987Z",
     "iopub.status.busy": "2021-06-08T10:27:44.656288Z",
     "iopub.status.idle": "2021-06-08T10:27:44.671212Z",
     "shell.execute_reply": "2021-06-08T10:27:44.671611Z",
     "shell.execute_reply.started": "2021-06-07T07:20:06.478142Z"
    },
    "papermill": {
     "duration": 0.032184,
     "end_time": "2021-06-08T10:27:44.671735",
     "exception": false,
     "start_time": "2021-06-08T10:27:44.639551",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [],
   "source": [
    "model.compile(optimizer='Adam',loss='categorical_crossentropy',metrics=['accuracy'])\n",
    "#model.summary()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "packed-iceland",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2021-06-08T10:27:44.696183Z",
     "iopub.status.busy": "2021-06-08T10:27:44.695639Z",
     "iopub.status.idle": "2021-06-08T10:27:45.221699Z",
     "shell.execute_reply": "2021-06-08T10:27:45.220662Z",
     "shell.execute_reply.started": "2021-06-07T07:20:06.505722Z"
    },
    "papermill": {
     "duration": 0.540319,
     "end_time": "2021-06-08T10:27:45.221861",
     "exception": false,
     "start_time": "2021-06-08T10:27:44.681542",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Found 1951 images belonging to 4 classes.\n",
      "Found 106 images belonging to 4 classes.\n"
     ]
    }
   ],
   "source": [
    "trdata = ImageDataGenerator(rescale = 1./255,\n",
    "                                   shear_range = 0.2,\n",
    "                                   zoom_range = 0.2,\n",
    "                                   horizontal_flip = True)\n",
    "\n",
    "tsdata = ImageDataGenerator(rescale = 1./255)\n",
    "traindata = trdata.flow_from_directory(directory=\"../input/cotton-disease-dataset/Cotton Disease/train\",target_size=(256,256),batch_size = 32,\n",
    "                                                 class_mode = 'categorical')\n",
    "testdata = tsdata.flow_from_directory(directory=\"../input/cotton-disease-dataset/Cotton Disease/test\", target_size=(256,256),batch_size = 32,\n",
    "                                                 class_mode = 'categorical')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "collected-disabled",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2021-06-08T10:27:45.254245Z",
     "iopub.status.busy": "2021-06-08T10:27:45.253636Z",
     "iopub.status.idle": "2021-06-08T10:40:34.573784Z",
     "shell.execute_reply": "2021-06-08T10:40:34.574348Z",
     "shell.execute_reply.started": "2021-06-07T07:20:06.837181Z"
    },
    "papermill": {
     "duration": 769.341224,
     "end_time": "2021-06-08T10:40:34.574503",
     "exception": false,
     "start_time": "2021-06-08T10:27:45.233279",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/20\n",
      "61/61 [==============================] - 61s 851ms/step - loss: 0.9600 - accuracy: 0.6804 - val_loss: 0.3171 - val_accuracy: 0.9151\n",
      "Epoch 2/20\n",
      "61/61 [==============================] - 38s 617ms/step - loss: 0.3720 - accuracy: 0.8851 - val_loss: 0.2725 - val_accuracy: 0.9151\n",
      "Epoch 3/20\n",
      "61/61 [==============================] - 37s 609ms/step - loss: 0.2854 - accuracy: 0.9095 - val_loss: 0.2024 - val_accuracy: 0.9623\n",
      "Epoch 4/20\n",
      "61/61 [==============================] - 37s 608ms/step - loss: 0.2818 - accuracy: 0.9044 - val_loss: 0.2071 - val_accuracy: 0.9245\n",
      "Epoch 5/20\n",
      "61/61 [==============================] - 37s 607ms/step - loss: 0.2245 - accuracy: 0.9241 - val_loss: 0.1855 - val_accuracy: 0.9528\n",
      "Epoch 6/20\n",
      "61/61 [==============================] - 37s 610ms/step - loss: 0.2474 - accuracy: 0.9222 - val_loss: 0.1180 - val_accuracy: 0.9623\n",
      "Epoch 7/20\n",
      "61/61 [==============================] - 37s 602ms/step - loss: 0.1949 - accuracy: 0.9285 - val_loss: 0.1724 - val_accuracy: 0.9528\n",
      "Epoch 8/20\n",
      "61/61 [==============================] - 37s 604ms/step - loss: 0.1682 - accuracy: 0.9401 - val_loss: 0.1111 - val_accuracy: 0.9528\n",
      "Epoch 9/20\n",
      "61/61 [==============================] - 37s 610ms/step - loss: 0.2057 - accuracy: 0.9236 - val_loss: 0.1058 - val_accuracy: 0.9717\n",
      "Epoch 10/20\n",
      "61/61 [==============================] - 37s 607ms/step - loss: 0.1565 - accuracy: 0.9442 - val_loss: 0.1196 - val_accuracy: 0.9717\n",
      "Epoch 11/20\n",
      "61/61 [==============================] - 38s 614ms/step - loss: 0.1494 - accuracy: 0.9416 - val_loss: 0.1207 - val_accuracy: 0.9623\n",
      "Epoch 12/20\n",
      "61/61 [==============================] - 37s 606ms/step - loss: 0.1589 - accuracy: 0.9430 - val_loss: 0.1548 - val_accuracy: 0.9528\n",
      "Epoch 13/20\n",
      "61/61 [==============================] - 37s 605ms/step - loss: 0.1617 - accuracy: 0.9430 - val_loss: 0.1512 - val_accuracy: 0.9623\n",
      "Epoch 14/20\n",
      "61/61 [==============================] - 37s 610ms/step - loss: 0.1493 - accuracy: 0.9452 - val_loss: 0.1156 - val_accuracy: 0.9717\n",
      "Epoch 15/20\n",
      "61/61 [==============================] - 37s 609ms/step - loss: 0.1778 - accuracy: 0.9393 - val_loss: 0.0996 - val_accuracy: 0.9717\n",
      "Epoch 16/20\n",
      "61/61 [==============================] - 37s 608ms/step - loss: 0.1453 - accuracy: 0.9468 - val_loss: 0.1049 - val_accuracy: 0.9717\n",
      "Epoch 17/20\n",
      "61/61 [==============================] - 38s 616ms/step - loss: 0.1443 - accuracy: 0.9532 - val_loss: 0.0783 - val_accuracy: 0.9717\n",
      "Epoch 18/20\n",
      "61/61 [==============================] - 37s 611ms/step - loss: 0.1460 - accuracy: 0.9447 - val_loss: 0.0522 - val_accuracy: 0.9811\n",
      "Epoch 19/20\n",
      "61/61 [==============================] - 38s 616ms/step - loss: 0.1235 - accuracy: 0.9540 - val_loss: 0.0763 - val_accuracy: 0.9623\n",
      "Epoch 20/20\n",
      "61/61 [==============================] - 37s 608ms/step - loss: 0.1241 - accuracy: 0.9570 - val_loss: 0.0788 - val_accuracy: 0.9717\n"
     ]
    }
   ],
   "source": [
    "from keras.callbacks import ModelCheckpoint, EarlyStopping\n",
    "checkpoint = ModelCheckpoint(\"densenet.h5\", monitor='val_acc', verbose=1, save_best_only=True, save_weights_only=False, mode='auto', period=1)\n",
    "early = EarlyStopping(monitor='val_acc', min_delta=0, patience=20, verbose=1, mode='auto')\n",
    "\n",
    "history = model.fit_generator( traindata,validation_data=testdata,epochs=20,callbacks=[checkpoint,early])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "subject-section",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2021-06-08T10:40:35.278198Z",
     "iopub.status.busy": "2021-06-08T10:40:35.277314Z",
     "iopub.status.idle": "2021-06-08T10:40:39.844063Z",
     "shell.execute_reply": "2021-06-08T10:40:39.844634Z",
     "shell.execute_reply.started": "2021-06-07T07:33:45.620235Z"
    },
    "papermill": {
     "duration": 4.944449,
     "end_time": "2021-06-08T10:40:39.844815",
     "exception": false,
     "start_time": "2021-06-08T10:40:34.900366",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "4/4 [==============================] - 4s 793ms/step - loss: 0.0788 - accuracy: 0.9717\n",
      "Test set accuracy: 97.16981053352356\n"
     ]
    }
   ],
   "source": [
    "accuracy = model.evaluate(testdata)\n",
    "\n",
    "print('Test set accuracy:', accuracy[1]*100)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "arctic-sender",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2021-06-08T10:40:40.564403Z",
     "iopub.status.busy": "2021-06-08T10:40:40.553336Z",
     "iopub.status.idle": "2021-06-08T10:40:40.811465Z",
     "shell.execute_reply": "2021-06-08T10:40:40.810871Z",
     "shell.execute_reply.started": "2021-06-07T07:33:50.247768Z"
    },
    "papermill": {
     "duration": 0.636958,
     "end_time": "2021-06-08T10:40:40.811613",
     "exception": false,
     "start_time": "2021-06-08T10:40:40.174655",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAY4AAAEWCAYAAABxMXBSAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjQuMSwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy/Z1A+gAAAACXBIWXMAAAsTAAALEwEAmpwYAABDN0lEQVR4nO3dd3zU9f3A8debhCRAAoEkzLBBBGQoEdwDRHGCG8RtpVpXa52tq1artrZa98+BqFVRUSx14QInCkE2CoQkQJgZrOz1/v3x+QaPkHEhd7mQvJ+Pxz1y9133vkty7/tsUVWMMcYYf7UIdQDGGGMOLJY4jDHG1IklDmOMMXViicMYY0ydWOIwxhhTJ5Y4jDHG1IklDmOqISK9RERFJNyPYy8XkW8bIi5jQs0Sh2kSRCRdRIpFJL7S9kXeh3+vEIVmTJNjicM0JWnApIoHIjIEaB26cBoHf0pMxtSFJQ7TlLwGXOrz+DLgVd8DRKSdiLwqIpkisk5E7hKRFt6+MBF5VESyRCQVOL2Kc18Skc0islFEHhCRMH8CE5F3RGSLiOwUka9FZLDPvlYi8k8vnp0i8q2ItPL2HSMi34vIDhHZICKXe9vnishvfK6xV1WZV8q6TkTWAGu8bf/2rrFLRBaKyLE+x4eJyJ9EZK2I7Pb2dxeRp0Xkn5VeyywR+YM/r9s0TZY4TFPyA9BWRAZ6H+gTgf9UOuZJoB3QBzgel2iu8PZdDZwBHAokAedVOncaUAr08445GfgN/vkY6A90BH4CXvfZ9ygwAjgK6ADcBpSLSE/vvCeBBGA4sNjP5wOYAIwCBnmPF3jX6AC8AbwjIlHevptxpbXTgLbAlUA+8AowySe5xgMneeeb5kpV7Wa3A/4GpOM+0O4CHgLGAZ8B4YACvYAwoBgY5HPeb4G53v0vgWt89p3snRsOdAKKgFY++ycBc7z7lwPf+hlrrHfddrgvbwXAsCqOuxOYWc015gK/8Xm81/N71x9dSxzbK54XWAWMr+a4n4Gx3v3rgY9C/fu2W2hvVvdpmprXgK+B3lSqpgLigZbAOp9t64Bu3v2uwIZK+yr09M7dLCIV21pUOr5KXunnQeB8XMmh3CeeSCAKWFvFqd2r2e6vvWITkVuAq3CvU3Eli4rOBDU91yvAxbhEfDHw73rEZJoAq6oyTYqqrsM1kp8GvFdpdxZQgksCFXoAG737m3EfoL77KmzAlTjiVTXWu7VV1cHU7iJgPK5E1A5X+gEQL6ZCoG8V522oZjtAHns3/Heu4pg9U1977Rm3ARcA7VU1FtjpxVDbc/0HGC8iw4CBwPvVHGeaCUscpim6CldNk+e7UVXLgLeBB0UkxmtDuJlf20HeBm4UkUQRaQ/c4XPuZuBT4J8i0lZEWohIXxE53o94YnBJJxv3Yf83n+uWA1OBf4lIV6+R+kgRicS1g5wkIheISLiIxInIcO/UxcA5ItJaRPp5r7m2GEqBTCBcRO7BlTgqvAj8VUT6izNUROK8GDNw7SOvAe+qaoEfr9k0YZY4TJOjqmtVNbma3Tfgvq2nAt/iGnmnevteAGYDS3AN2JVLLJcCEcBKXPvADKCLHyG9iqv22uid+0Ol/bcAy3AfzjnAI0ALVV2PKzn90du+GBjmnfMYrr1mK64q6XVqNhv4BFjtxVLI3lVZ/8Ilzk+BXcBLQCuf/a8AQ3DJwzRzomoLORljaiYix+FKZj3VPjSaPStxGGNqJCItgZuAFy1pGLDEYYypgYgMBHbgquQeD2kwptGwqipjjDF1YiUOY4wxddIsBgDGx8drr169Qh2GMcYcUBYuXJilqgmVtzeLxNGrVy+Sk6vrnWmMMaYqIrKuqu1WVWWMMaZOLHEYY4ypE0scxhhj6qRZtHFUpaSkhIyMDAoLC0MdStBFRUWRmJhIy5YtQx2KMaYJaLaJIyMjg5iYGHr16oXPNNlNjqqSnZ1NRkYGvXv3DnU4xpgmoNlWVRUWFhIXF9ekkwaAiBAXF9csSlbGmIbRbBMH0OSTRoXm8jqNMQ2jWScOY4wJmuy1kDwVykpDHUnANds2jlDLzs5mzJgxAGzZsoWwsDASEtwAzfnz5xMREVHtucnJybz66qs88cQTDRKrMaYOinbD14/CvKehvAS2/QKn/T3UUQWUJY4QiYuLY/HixQDcd999REdHc8stt+zZX1paSnh41b+epKQkkpKSGiJMY4y/VGHZO/DZPbB7Mwy7CFq2gvn/B3H9YNSUUEcYMJY4GpHLL7+cqKgoFi1axNFHH83EiRO56aabKCwspFWrVrz88ssMGDCAuXPn8uijj/LBBx9w3333sX79elJTU1m/fj2///3vufHGG0P9UoxpXjYvgY9ugw0/QJfhcMGr0H0klJe5JPLJ7dC+Fxx0cqgjDYigJg4RGQf8GwjDLQLzcKX9PXHLdibglsa8WFUzRORE3NKYFQ4GJqrq+yIyDTge2Ontu1xVF9cnzr/8bwUrN+2qzyX2MahrW+49c3Cdz8vIyOD7778nLCyMXbt28c033xAeHs7nn3/On/70J9599919zvnll1+YM2cOu3fvZsCAAVx77bU2ZsOYhpCfA1/+FZJfhtZxcNaTMPxiaOE1H7cIg3NegJdPhRlXwJWzofMhoY05AIKWOEQkDHgaGAtkAAtEZJaqrvQ57FHgVVV9RURGAw8Bl6jqHGC4d50OQApuLeQKt6rqjGDFHkrnn38+YWFhAOzcuZPLLruMNWvWICKUlJRUec7pp59OZGQkkZGRdOzYka1bt5KYmNiQYRvTvJSVwsKX4csHXJvGqN/CCXdCq9h9j42MhoveghdGwxsXwtVfQkynBg85kIJZ4hgJpKhqKoCITAfGA76JYxBws3d/DvB+Fdc5D/hYVfODFej+lAyCpU2bNnvu33333Zx44onMnDmT9PR0TjjhhCrPiYyM3HM/LCyM0tKm14vDmEYj/Tv4+HbYugx6HQun/h06Dar5nLZdYdJ0V/KYPgku+wAiWjdMvEEQzO643YANPo8zvG2+lgDnePfPBmJEJK7SMROBNytte1BElorIYyISSRVEZIqIJItIcmZm5v69ghDbuXMn3bq5t2zatGmhDcaY5m7nRphxJUw7DQp3wPmvwGX/qz1pVOg6HM59ETb+BO9fA+XlwYw2qEI9juMW4HgRWYRrt9gIlFXsFJEuwBBgts85d+LaPA4HOgC3V3VhVX1eVZNUNamim+uB5rbbbuPOO+/k0EMPtVKEMaFSUui61z6VBD9/AMffDtfNh8EToK6Daw8+HU7+K6z8L8x5ICjhNoSgrTkuIkcC96nqKd7jOwFU9aFqjo8GflHVRJ9tNwGDVbXKfmwicgJwi6qeUVMsSUlJWnkhp59//pmBAwf6/XoOdM3t9QZMXjZoPb4ZRka7LpnNkaprPK7P+xdqGQtg9p9gexocfAac8qDrHVUfqvDB72HhNBj/DBw6OQCBViNrDcT33+/TRWShqu7T9z+YbRwLgP4i0htXkpgIXFQpqHggR1XLcSWJqZWuMcnb7ntOF1XdLG4ejQnA8uCEb5o1VXj/WlhSuZa0jiJi4ITbYeRvIbz6QZ1NztaV8PFtkP5NqCOpv/iD4OL3oN+YwFxPBE57FLanw/9ugvY9odcxgbl2he3rXMJb9RFM+Qq6DA3o5YOWOFS1VESux1UzhQFTVXWFiNwPJKvqLOAE4CERUeBr4LqK80WkF9Ad+KrSpV8XkQRAgMXANcF6DaYZm/uwSxpJV0HHepTU1nwGn94FP70K4x4O3IdPY1Ww3b1381+AqLZw4p+hVftQR7X/WrWHgWcFPumHtXRtJC+NhemT4TdfQHy/+l+3OB++exy++zdIC/f+xx9U/+tWErSqqsbEqqqa3+utl6Vvw3tXu/7445+qez12Zas/dQPAclJhwOmuuqNDE5vivrwMFv0HvviLSx4jroDRd0HrDqGOrHHLSYMXx0BULPzm8/1/v1Th51kw+8+wcwMcci6MvR/a1a9bfnVVVaFuHDemcVn/A/z3OtfN8ozH6p80wI0W/t0PcNJ9kDoXnh4FXz7ovh02BRsWuDEK/7vRfbud8hWc8S9LGv7o0BsmvuE+7N+6BEqL636NbT/Dq+Ph7Ushsi1c/iGcN7XeSaMmljiMqZCTCtMvgnbd3ZQRgayeCI+EY/4ANyTDoPHw9d/hqcNh+Xvu2+KBaPdWmHktvHQS5G6Fc1+CKz4OeH16k9fjCNdIvu5b1+bh799DwQ745E549mg35clpj8Jvvw58e0kVbK4qY8D9E75xoesBNPmd4H1bbtsVzn0Bkq6Ej29101AkT4VTH4FOjWcgao1Ki+HH5+Crv0NZkUuIx97iepCZ/TP0fMhZC3Mfcm0dx/6x+mPLy2Hx665aMC8LRlwOo++GNpWHwAWPJY4Qqc+06gBz584lIiKCo446KuixNnllJa6Yn5MGl74PcX2D/5w9j3RVOgunubmOnjsWDv8NnHhn425MTvnCjZrOXgP9T4FxDzXM+9UcHH87ZKfAF/dDh75unEhlGcnw0a2w6SfoPgomz3ADCxuYJY4QqW1a9drMnTuX6OhoSxz1pQof/hHSvoIJzzZIMX+PFmFw+FUw+GyY8yAseAGWz4Ax98Chl7j9jUVOmmt4XfUhdOgDF70NB50S6qiaFhE46ynYsR5m/tZVmSaOcPtyt8Hn97mSRnRnOPt5GHpBYNrg9oO1cTQiCxcu5Pjjj2fEiBGccsopbN68GYAnnniCQYMGMXToUCZOnEh6ejrPPfccjz32GMOHD+ebb5pAX/lQmfcU/PSKqxoYflHtxwdD6w5w+j9d/XT8AFfP/cJo2DA/NPH4Ks53E/k9Pco17J90n2vot6QRHC2jXGN5dCd4c6Jrd5v3NDw5wvX2O/om10427MKQJQ2wEofz8R2wZVlgr9l5CJz6cO3HeVSVG264gf/+978kJCTw1ltv8ec7bmXqM4/z8MMPk5aWRmRkJDt27CA2NpZrrrmmzqWUBlVeBoteg91bYOSUxtnD5ucP4NO7YdAEOPGuUEfj/mau+AiWv+viemksDJ0IY/8CMZ0bNhZVWPk+zL4LdmXAkPNd9862XRs2juaoTbxrZ3txLDyZBFoG/ca6cUCBGOsRAJY4GomioiKWL1/O2LFjASgrK6NLXFvYmcHQIUOYPHkyEyZMYMKECaEN1B++s4cC/Ph/rvrlsEsbT/XLpsVurEa3EXD2c7+unxBqIjDkPDhoHHzzT1ci+uVDOP42GHVNw4w+37rC/f7Sv3HJ7NwXXZuMaTgJA2Dif1wHhCOvc38PISxhVGaJA+pUMggWVWXw4MHMmzfPbSjOh6xVgPLh9Jf4evEa/ve///Hggw+ybFmAS0eBsnMjfHa3+8bcNhHOn+aWzPz4dm9unpfh1H9Aj1Ghj/PNiW7hnUlvNs65pCKj4aR74dCL3dQRn93tRp+f+jD0Oyk4z1mwHeY8BAtedKO+T/+X67HTWJJ9c9P7OHdrhBrJ1ywTGRlJZmbmnsRRsnsbK1alUh7Wig0pyznxhON55JFH2LlzJ7m5ucTExLB79+4QR+0pLXLfjitmDz3uNrh+gWv07TzEDUg69yXIzYSpJ8N7U1wVVigU5cKbF7qfF70F0R1DE4e/4vq6OC96x3UV/s+58OYk11gdKOVlrnfXkyNcA33SFXDDT67h3pKGqYIljkaiRYsWzJgxg9tvv51hw4Yx/KiT+H7xL5S1TuDi6+5kyCGHcOihh3LjjTcSGxvLmWeeycyZM0PfOL7qE9dw+sX90Hc0XD8fRv9570VqKqpfrl/gGqFXzHQfUt/9e/9Gyu6v8jJXPbV1hSsNHSjjJsAbfT4PTvoLpH3tjT5/AIrz6nfdDfO9Ud83uYb5KV+5hvrG2CZlGg2bq6oxKs6DrNUQ2wNadXBTCrQIh4T9n6ws4K83KwVm3wlrPnXTTNRlAr/sta5r5+qPXVXWuEegf5CqX3zN/rNrMzjtURh5dfCfL1h2bYbP74Wlb7kqwZP/6kp3dakD373Fde9c8ibEdHXXOOTcRlWPbkLP5qo6kBRsBwSi2rl/5DbxUJLXOOY2KtoNn90DzxwB6+bByQ/CNd/VbdbXuL5w0XQ3eEkVXq+ofkkNXtzJU13SGHXNgZ00ANp2gXOehytnQ+v2bvT5K2e6klRtSovhuydcb53l78IxN7uS4JDzLGkYv1niaGxU3fQXkTGulAGu2kBaQF4Il8BVdf3In0xyVUxDL4AbFsJR1+9/T5/+Y73J/yqqX46AL/5a/+qXytZ+CR/eAv1PhlP+Fthrh1KPI7wJBR+DrcvhuWPcqOKC7VUfn/I5PHuUa2jvdbT33t9rU4WYOmvWiaNRVtMV50F5yd7TTrQId1VWBdvd9Bh1VO/XuXkJTB3n2gfadoGrPocJz0BMp/pdF1zSOeb3cH2ym2Lhm0e9yf/eDczkf9t+gbcvg4SD3YyhTa2xt0WYm/fqhp/c2iELXnTtRwunuTYdcA3pb17kGta13DW0X/SWTRVi9luzbeNIS0sjJiaGuLg4pDEV0XdsgPxs1xvJ90OupAAyf4GYLnUaDKaqZGdns3v3bnr3ruMaEPk5bh6l5Jdd19WT7nVrVARzzMP6H9y35i1LoecxbgK9/e0uW14Cs25wvb5+8wXEdg9srI3RlmWu+/O676DLMDc9/PwX3JeP42+FI37nZuo1NSosKSNjewHFpeUM7BITks+I4tJyVm7eRdfYKBKiI0MSQ3VtHEFNHCIyDvg3bgXAF1X14Ur7e+KWi00AcoCLVTXD21cGVAxYWK+qZ3nbewPTgThgIXCJqtbYNaeqxFFSUkJGRgaFhYX1e5GBpAq7Nrl/7Dbx++7P3QblpS551OGPKCoqisTERFq2bFm3WF49yw3mGzkFTrgDWsX6f359lJe5aUC++CsU5NTvWuGt4IoP3UC/5kL119HnuzfBkAu8Ud9dQh1Zo1FSVs6mHQVkbC9gQ04+G7bn+9wvIHN30Z5jk3q25+aTD+KovlX8TwYptncXZvDklyls3FEAQNuocPp1jKZvQjT9Ov56S2zfmrAWwUsoDZ44RCQMWA2MBTJwa5BPUtWVPse8A3ygqq+IyGjgClW9xNuXq6r7VL6KyNvAe6o6XUSeA5ao6rM1xVJV4miUUr9yH9bnT3O9ZCr75UO3XsQFr7o1HYKp4rlO/QeMmhLc56pOwQ5XTVYfHfo0j5JGVYrzXem1mb7+snJl0frtrMt2iWFDTgEZXoLYvLOAcp+PvrAWQtfYKLq3b01i+1Z0b9+a7h1asyO/mGe/WsvWXUUc2SeOP558EEm9gtNVuaxceX/RRp74cg3rsvMZltiOy4/uxY78ElK25ZKyLZe1mblk5f76PTkyvAW949vslUz6dYymd3wbIsPrXy0bisRxJHCfqp7iPb4TQFUf8jlmBTBOVTeIK4ftVNW23r59Eod3TCbQ2VvTfK/nqM4Bkzj+dxMsfQduTdl7HESF8jL493DXTfeKD4MXR2kxPDMKWrSEa7+HMJtgIBRKysrZnldMx7ZRoQ7lgJNfXMrvXv+JuatchxIR6BQTRfcOrfYkh8QOrb0E0YrObaMID6u6CrawpIw3flzPM3PXkpVbxHEHJXDz2IMY3j02ILGWlysfLNvM45+vJjUzj0Fd2nLz2IMYM7BjldVTO/KL90okKdtyScnMJWN7wZ5mwRYCPTq0pl/HaO45YzA94qr4PPFDdYkjmJ8I3YANPo8zgMpzTSwBzsFVZ50NxIhInKpmA1EikgyUAg+r6vu46qkdqlrqc81uVT25iEwBpgD06NEjIC8oqMpKYOUsGDCu6qQBv07D/fm9rutlsAawLXjBdY2dPMOSRohsyMnnd6//xPJNOzntkC787sS+DO7aLtRhHRCyc4u4ctoClm3cyd1nDGL0wR3pGhu139/Ao1qGceUxvZk4sjuvzVvHc1+tZcLT33HSwI78YexB+/17KS9XZq/YwmOfr2b11lwO6hTNs5MP45TBnWlRQ/VTbOsIknp12KfkU1BcRmqWl1C8ZLJ2Wx5REYFvkwz1p8ItwFMicjnwNbAR8LqC0FNVN4pIH+BLEVkG7PT3wqr6PPA8uBJHQKMOhrSvXH3+4HNqPu6wS90qYfOfhzP/Hfg48nPgq0eg7xjXXdY0uC9+3srNby+hXJWLRvZg1uJNfLhsMycOSOD60f0Y0dNGdVdnXXYel02dz5ZdhTx/SRInDQpAzz9P64hwfnt8XyYf0ZNp36Xx/NepnP7Et4wb3Jk/jD2IAZ1j/LqOqvLFz9v412erWbl5F30S2vDEpEM5Y0iXGhNGbVpFhDG4a7sG+YIRzMSxEfCtXE30tu2hqptwJQ5EJBo4V1V3ePs2ej9TRWQucCjwLhArIuFeqWOfax6wls90C83XNoFd6w5uiuulb7u1EQK9Wtzch9w8Tqc8GNjrmlqVlpXzr89W88zctQzu2pZnJ4+gR1xrbht3MK/NS+elb9M499l5HNGnA9ef2J+j+zWyHoEhtixjJ1dMm09pufL6b45gRM/grKQYHRnO9aP7c8mRvXjp2zSmfpvG7JVbOGNoV35/Un/6JlQ9LkZV+XpNFv/6bDVLNuygR4fW/PP8YYwf3rXaarLGKphtHOG4xvExuA/3BcBFqrrC55h4IEdVy0XkQaBMVe8RkfZAvqoWecfMA8ar6kqvQf1dn8bxpar6TE2xNPo2jtJieLQfHHQqnPN/tR+/ZZkb7HXyA3DUDYGLI3MVPHOkm+Tu9H8G7rqmVpm7i7jxzUXMS81m0sju3HvmYKJa7l21kl9cyhs/rueFb1LZuquIYd1jue6Evpw0sFO9vqk2BV+tzuTa/yykfesIXrlyJP06Ntygxu15xTz/TSrTvkunqLSMCYd246Yx/ekZ12bPMd+vzeJfn64med12usW24obR/Th3RCItG3nCCFV33NOAx3Hdcaeq6oMicj+QrKqzROQ84CFAcVVV13nJ4ijg/4By3CDFx1X1Je+afXDdcTsAi3BdeIuoQaNPHKs+cTO21mU5zqmnwq6NcOOiwA1q+895btK7Gxc16ML3zd38tByuf+MndhWW8MCEIZw3IrHG44tKy3h34Uae+2ot63PyGdApht+d2JfTh3Q54L65BsJ7P2Vw24yl9O8Uw7QrDqdTiDoTZOUW8X9freXVeesoLVfOH5HISQM78dK3acxLzaZT20iuP7EfFxzePSA9nhpCSBJHY9HoE8d7U2D1bLhljf/Tdyx/z81RNGk6DDi1/jGkfO5GFge6FGOqpaq88E0qj3yyih4dWvPsxYdxcOe2fp9fWlbOB0s38/ScFNZsy6VnXGuuPb4v5xyWSER4/ROIqlJaro32W7Gq8txXqTzyyS8c1TeO5y4ZQduoOoxVCpJtuwp5Zu5a3vhxPcVl5cRHR3DtCf2YPKrHPqXIxs4SR2NNHCWF8I9+MHg8jH/a//PKSuDxodDxYLhkZv1iKCuF5452I6yv+9FGFjeAnQUl3PrOEj5duZXThnTmkXOHErOfH3rl5cqnK7fy9JwUlm3cSZd2UVx9bB8mjexBq4h9P6iKS8vJyi0ic3cR23ZX/Czc83jb7iKyvO0l5eV0b9/61zECCdH09e63axW6D+mycuWvH6xk2vfpnDWsK4+ePywgyTKQNu0oYNH6HZx4cAKtI0LdD2n/hKI7rvFHymdQvLv23lSVhbV0cxTNeQCy1kB8//2P4adpbjqTC/9jSaMBLN+4k9+9/hObdhRw9xmDuPLoXvVq5G7RQhh3SGdOGdyJb9Zk8dScFO7/YCVPz0nh9KFdyC0s3StBbM+ver6zuDYRJMREkhATSd+ENnSMiSIiTFiblcfabbl8m5JFcWn5nuMTYiLp541k7pvQhn4dY+jXMZpObYM7PUZhSRk3v72Yj5Zt4epje3PnqQMbZRtP19hWdI1thKtLBoCVOELtnStcV9w/rq77mIncTHhskFve87R/7N/zF+yAJw+DhIFw+Qc2tXYQqSpvJ2/g7v+uoEPrCJ6efGjQutYuSM/hqS9TmJ+WQ4c2EXRsG0lCdKT3M2qvxx1jooiLjqi1SqqsXNmQk79nwNmeQWjbctldVLrnuJjIcPp4pZOBXWI4tn8CB3WKDkgy2VlQwtWvJjM/LYe7Th/Ib47tU+9rmupZiaMxKs6D1Z/A0Av3b6BddIIrqSx+A0bf7daJrqtvHnVjN8b9zZJGEBUUl3H3f5czY2EGx/aP5/ELhxMXHbzS3eG9OvDKlSMDes2wFkKv+Db0im/DSfw6PkJV2ba7aM+gs4qE8s2aTN79KQP4ma7tojh+QEdOHJDA0f3iaRNZ97/3zTsLuHzqAlKzcvn3xOGMH17l2F/TACxxhNLqT6AkHw6pYzWVr1FTYOl0t5LbqN/W7dycVPjhORg+2c2kaoIiNTOX373+E6u27uamMf25cUz/oE5M19BEhE5to+jUNoqj+u09EeDmnQXMXZXJ3FXbmLV4I2/OX0/LMGFk7w6cOKAjJwxIoG9C7aWR1Vt3c9nU+ewuLGXaFSM5utLzmIZlVVWhNH0yZCyAm3+uX5faF0ZD4U64bkHdpjx/62JI+RJu/KlOU7Ub/320bDO3zVhKeJjw+IXDOWFAx1CHFDLFpeUkp+cwd3Umc37ZxpptuQAktm+1J4kc2Tdun4bkBek5XDVtAZEtw5h2xeE29UoDsqqqxqZwF6z5DEZcVv9xGCN/CzOnQOoc/5dwTfsGfv4fnHiXJY0AU1Xmrc3m1Xnr+GTFFoZ1j+WZyYfRrYk2lPorIrwFR/WL56h+8fzptIFkbM/fUxqZsTCD135YR0R4C47oE8cJByVw4sEdWbVlFzdOX0xi+1a8csVIunfYv8n6TGBZiSNUlrzlPuyvnO2WAK2P0iJ4bLBbc+Kit2o/vrwMnj/BtW3ckLz/CyWZvewqLOE97wNwbWYesa1bcskRPblhdP9G11W0sSkqLWN+Wg5zV2UyZ9U2UjN/XT740B6xvHTZ4XRos59LFJv9ZiWOxmbFe9C2GyQGoAEzPBJGXAFf/8MtE9qhlpX+lrzpVtg750VLGgHw8+ZdvPbDOt5ftJH84jKGdY/l0fOHccbQLgfcgK9QiQwP49j+CRzbP4G7zxjE+ux85q7exo78Eq4+tk+V41FM6FjiCIWC7ZDyhWvMDtQyrElXwrf/cmtO1zRBYVEufHE/dEuCIecF5rmboeLScj5evpn//LCOBenbiQxvwVnDunLJkT0Zmhgb6vAOeD3iWnPpkb1CHYaphiWOUPjlQ7cedl0H/dWkbRcYeCYseg1O/BNEtKn6uO8eh9ytbrBfE+5+q6qkZuWxdWch3dq7gViBmDpj444C3vxxPdMXrCcrt5ieca256/SBnDcikdjWVpVimgdLHKGw/D2I7QndDgvsdUf+FlbMdFOuJ12x7/4dG+D7J+GQ86B7YPv4NwYbcvKZtzab79dmMS81m627fp37soVAl3at3Mpv3qpvFcuDJrZvRae2UdV2kS0vV75bm8Vr89bx+c9bUWDMwR255MheHNsvvlGOWjYmmCxxNLS8bEid6yYSDPQ3/h5HQOchbpGnEZfve/0v/uJ+nnRfYJ83RDbvLGDe2mwvWWSzcUcBAPHRERzZN54j+8TRM641G3cUkLG9gIwct/b0dylZbN1diG+/kJZhQrfYX5NKopdUtu0q5I0f15OalUdcmwiuOb4vk0b2sN49plmzxNHQfp4FWla/QX/VEXGljlnXQ/q30PvYX/dtWADL3oFjb4HY7tVfoxHLyi1yiSLVJYu0LNfzJrZ1S47oHceU4/pwVN84+nWsfUBZUWkZm3YUssFLJhnbC7z7BXy6YivZecV7jj2sRyyPXzicU4d0PmCmwzYmmCxxNLQV70GHvtB5aHCuP+Q8+OwemP9/vyYOVZh9J0R3gmP+EJznDYJdhSV8n5LND6mu+mn1VjdgLDoynFG9OzB5VA+O7BvHwM5t61xdFBkeRu/4NvSOr7otKL+4lIztBbQQadBFgYw5EFjiaEi521xJ4Ng/Bq9humUrty7590+4No3Y7rD8XTdC/aynILJxfwhm5Rbx+cqtfLJiC9+lZFFSprRqGUZSr/acfWgiR/aN45CubYO+YFHriHAO6uTfGtLGNDdBTRwiMg74N24FwBdV9eFK+3sCU4EEIAe3ml+GiAwHngXaAmXAg6r6lnfONOB4YKd3mctVdXEwX0fArPwvaHlge1NV5fCrXOJIngrH3waf3evaPoZfFNzn3U8bdxQwe/kWPlmxheT0HMoVenRozRVH9+akgZ0Y3j3WBtAZ04gELXGISBjwNDAWyAAWiMgsVV3pc9ijwKuq+oqIjMYtI3sJkA9cqqprRKQrsFBEZqvqDu+8W1V1RrBiD5rl70HCwdBpUHCfJ7YHDDgNfnrFTWeyKwPOfi5wS8wGQMq2XGav2MLsFVtYmuG+AwzoFMP1o/szbnBnBnaJCeqaDsaY/RfMEsdIIEVVUwFEZDowHvBNHIOAm737c4D3AVR1dcUBqrpJRLbhSiU7ghhvcO3aBOvnwQl3NszzjZwCv3zgRpMffMbeDeUhoKqs2LSLT7ySRYo3wd3w7rHccerBnDK4c7XtDcaYxiWYiaMbsMHncQYwqtIxS4BzcNVZZwMxIhKnqtkVB4jISCACWOtz3oMicg/wBXCHqhZRiYhMAaYA9OjRo/6vpr5WvA9ocHpTVaX3cW5xpuwUGHt/wzxnJUWlZSzN2OmSxfItbNxRQFgLYVTvDlxyRE9OHtyJLu1syhNjDjShbhy/BXhKRC4HvgY24to0ABCRLsBrwGWqWrFm5Z3AFlwyeR64Hdjnk1FVn/f2k5SUFPqZHJe/C52G1G+J17oQgbOfhd1bIa5vwC6rquwsKKl2vWrfnzsL3BKlEWEtOLZ/PDed1J+TBnayyeqMOcAFM3FsBHwHDCR62/ZQ1U24EgciEg2cW9GOISJtgQ+BP6vqDz7nbPbuFonIy7jk07htXwcbk2HMPQ37vF0P3e9TS8rK+WT5Fn5Izd4rGWTmFu217nSFqJYt6BgTRUJMJP07RnNU3zg6xkTSOz6a4wckEL0fK74ZYxqnYP43LwD6i0hvXMKYCOzVrUdE4oEcrzRxJ66HFSISAczENZzPqHROF1XdLK7ldAKwPIivITBWzHQ/g92bKgB2FZbw1vwNvPxdGpt2FtI2Kpwu7VrRsW0kfeLbkOCtUZ0QE0nHmMg9P6Mjw60x25hmImiJQ1VLReR6YDauO+5UVV0hIvcDyao6CzgBeEhEFFdVdZ13+gXAcUCcV40Fv3a7fV1EEgABFgPXBOs1BMyK99y3/9qmOw+hDTn5TPs+nbcWbCC3qJQj+nTg/vGHMPrgjjYXkzFmL7aQU7Blr4UnD4OTH3DzUzUyi9Zv58Vv0/h42WZaiHDG0C5cdUwfhiTa8pzGNHe2kFOorHjP/Rx8dmjj8FFWrny2cgsvfpNG8rrtxESFc/Vxfbj8qF7Wy8kYUytLHMG2fCZ0HwXtEkMdCXlFpbyTvIGp36WzPief7h1ace+Zgzg/qbs1Xhtj/GafFsGUuQq2rYBxj4Q0jC07C5n2fTpv/LiOXYWlHNYjljtPPZiTB3eudg0KY4ypjiWOmiS/DFmraz+uOluXAwKDxgcspLpI2ZbL03NS+N+STZSrMu6Qzlx1TB9G9GwfkniMMU2DJY6apM6BtXPqd40h57tlXRtQaVk5L3yTxmOfraZlmHDJkT258ujetviQMSYgLHHU5IJXQx1BnaVsy+WWd5aweMMOTj2kM3+dcAjx0ZGhDssY04RY4mgiysqVqd+m8Y9PV9E6IownJh3KmUO72KA8Y0zAWeJoAlIzc7l1xlIWrtvO2EGdePDsQ+gYExXqsIwxTZQljgNYebny8vfp/P2TX4hqGcZjFw5jwvBuVsowxgSVJY4DVHpWHrfNWMr89BzGHNyRv50zhE5trZRhjAk+SxwHmPJy5dV56Tz8yS+0DGvBo+cP49zDrJRhjGk4ljgOIOuz87l1xhJ+TMvhhAEJPHzOUDq3s1KGMaZhWeI4AJSXK6/PX89DH/1MmAh/P3co5yclWinDGBMSljgauQ05+dz+7lK+X5vNsf3jefjcoXSLtYkIjTGhY4mjkVJV3py/gQc/XAnA384ewqSR3a2UYYwJOUscjdDGHQXc8e5SvlmTxVF94/j7eUNJbG/ThRhjGocWwby4iIwTkVUikiIid1Sxv6eIfCEiS0Vkrogk+uy7TETWeLfLfLaPEJFl3jWfkCb0FVxVeWvBek557GsWrtvOXyccwn+uGmVJwxjTqNSaOETkTBGpc4IRkTDgaeBUYBAwSUQGVTrsUdy64kOB+4GHvHM7APcCo4CRwL0iUjGl67PA1UB/7zaurrE1Rpt3FnD5ywu4/d1lHNKtLbN/fxyXHNHTlm01xjQ6/iSEC4E1IvJ3ETm4DtceCaSoaqqqFgPTgcrziw8CvvTuz/HZfwrwmarmqOp24DNgnIh0Adqq6g/q1rx9FZhQh5gaHVVlxsIMTn7sa+an5fCXswbzxm+OsJlsjTGNVq1tHKp6sYi0BSYB00REgZeBN1V1dw2ndgM2+DzOwJUgfC0BzgH+DZwNxIhIXDXndvNuGVVs34eITAGmAPTo0aOmlxgyW3cV8qf3lvHFL9s4vFd7/nHeMHrFtwl1WMYYUyO/qqBUdRcwA1dq6IL7kP9JRG6o5/PfAhwvIouA44GNQFk9rwmAqj6vqkmqmpSQkBCISwaMqjJzkStlfLc2i7vPGMRbU460pGGMOSDUWuIQkbOAK4B+uKqhkaq6TURaAyuBJ6s5dSPQ3edxordtD1XdhCtxICLRwLmqukNENgInVDp3rnd+YqXte12zsdu2u5A/z1zOZyu3MqJne/5x3lD6JESHOixjjPGbP91xzwUeU9WvfTeqar6IXFXDeQuA/iLSG/fhPhG4yPcAEYkHclS1HLgTmOrtmg38zadB/GTgTlXNEZFdInIE8CNwKdUnrkZFVZm1ZBP3zlpBfnEZfz5tIFce09vW/DbGHHD8SRz3AZsrHohIK6CTqqar6hfVnaSqpSJyPS4JhAFTVXWFiNwPJKvqLFyp4iGv3eRr4Drv3BwR+Ssu+QDcr6o53v3fAdOAVsDH3q1Ry8ot4q6Zy/lkxRaGd4/l0fOH0a+jlTKMMQcmcZ2TajhAJBk4yusZhYhEAN+p6uENEF9AJCUlaXJyckie+8Olm7n7v8vJLSzl5pMP4jfH9CY8LKjDZ4wxJiBEZKGqJlXe7k+JI7wiaQCoarGXPEwtnvpyDY9+upqhie345/nD6N8pJtQhGWNMvfnz1TfTayAHQETGA1nBC6np+HTlVg7tEct71x5lScMY02T4U+K4BnhdRJ4CBDe+4tKgRtUEqCppWXlMGN7NqqaMMU2KPwMA1wJHeN1lUdXcoEfVBOTkFbO7sNTGZhhjmhy/ZscVkdOBwUBUxZyCqnp/EOM64KVn5wHQO96mDjHGNC3+THL4HG6+qhtwVVXnAz2DHNcBLzXTJY5ecVbiMMY0Lf5Uvh+lqpcC21X1L8CRwEHBDevAl56dR1gLsckKjTFNjj+Jo9D7mS8iXYES3HxVpgbpWfl0b9+KltYwboxpYvxp4/ifiMQC/wB+AhR4IZhBNQVpWXnWMG6MaZJqTBzeAk5fqOoO4F0R+QCIUtWdDRHcgUpVSc/OY1SfDqEOxRhjAq7GehRv8sGnfR4XWdKo3bbdReQXl9HbShzGmCbInwr4L0Tk3Ka0tnewpWVZjypjTNPlT+L4LfAOUORNab5bRHYFOa4DWnpWxRgOSxzGmKbHn5HjNslSHaVl5RER1oKusa1CHYoxxgScPysAHlfV9soLO5lfpWXl0SOutS3SZIxpkvzpjnurz/0oYCSwEBgdlIiagPTsPGvfMMY0WbW2cajqmT63scAhwHZ/Li4i40RklYikiMgdVezvISJzRGSRiCwVkdO87ZNFZLHPrVxEhnv75nrXrNjXsU6vOMjKy5X07Hybo8oY02T5NclhJRnAwNoOEpEwXFfesd45C0Rklqqu9DnsLuBtVX1WRAYBHwG9VPV14HXvOkOA91V1sc95k1U1NEv61WLTzgKKS8vpHW9LwxpjmiZ/2jiexI0WB1dCGY4bQV6bkUCKqqZ615kOjAd8E4cCbb377YBNVVxnEjDdj+drFNKz8gHoZSUOY0wT5U+Jw/ebfSnwpqp+58d53XCLPlXIAEZVOuY+4FMRuQFoA5xUxXUuxCUcXy+LSBnwLvCAVrFwuohMAaYA9OjRw49wAyMt27riGmOaNn8SxwygUFXLwFVBiUhrVc0PwPNPAqap6j9F5EjgNRE5xBuxjoiMAvJVdbnPOZNVdaOIxOASxyXAq5UvrKrPA88DJCUl7ZNYgiUtM49WLcPoFBPVUE9pjDENyq+R44DvgIRWwOd+nLcR6O7zONHb5usq4G0AVZ2H67UV77N/IvCm7wmqutH7uRt4A1cl1mikZ+fRM641LawrrjGmifIncUT5Lhfr3fenAn8B0F9EeotIBC4JzKp0zHpgDICIDMQljkzvcQvgAnzaN0QkXETivfstgTOA5TQi6Vl5Vk1ljGnS/EkceSJyWMUDERkBFNR2kqqWAtcDs4Gfcb2nVojI/SJylnfYH4GrRWQJrmRxuU97xXHAhorGdU8kMFtElgKLcSWYRjPFe2lZOetz8m06dWNMk+ZPG8fvgXdEZBNu6djOuAbrWqnqR7gutr7b7vG5vxI4uppz5wJHVNqWB4zw57lDIWN7AaXlaiUOY0yT5s9cVQtE5GBggLdplaqWBDesA5P1qDLGNAe1VlWJyHVAG1Vd7vVuihaR3wU/tANPuk2nboxpBvxp47jaWwEQAFXdDlwdtIgOYOlZeURHhhMfHRHqUIwxJmj8SRxhvos4eVOJ2CdjFVK9HlW25pUxpinzJ3F8ArwlImNEZAyu99PHwQ3rwJSenWc9qowxTZ4/ieN24EvgGu+2jL0HBBqgqLSMjdsL6B1nc1QZY5o2f6ZVLwd+BNJxo7RH48ZlGB8bcvIpV+idYCUOY0zTVm13XBE5CDeX1CQgC3gLQFVPbJjQDixpFbPiWo8qY0wTV9M4jl+Ab4AzVDUFQET+0CBRHYAquuLaGA5jTFNXU1XVOcBmYI6IvOA1jFt3oWqkZecR27olsa2tw5kxpmmrNnGo6vuqOhE4GJiDm3qko4g8KyInN1B8B4y0TJvc0BjTPPjTOJ6nqm+o6pm4qdEX4XpaGR/p2Xn0tvYNY0wz4E933D1UdbuqPq+qY4IV0IGooLiMzTsLbQyHMaZZqFPiMFVbl+PNUWWJwxjTDFjiCIC0TJc4+ljiMMY0A5Y4AqBiOnUrcRhjmoOgJg4RGSciq0QkRUTuqGJ/DxGZIyKLRGSpiJzmbe8lIgUisti7PedzzggRWeZd8wlpBDMKpmflER8dSXSkP+tiGWPMgS1oicObRfdp4FRgEDBJRAZVOuwu3JKyh+LWJH/GZ99aVR3u3a7x2f4sblr3/t5tXLBeg7/SsvKsmsoY02wEs8QxEkhR1VRVLQamA+MrHaNAW+9+O2BTTRcUkS5AW1X9wVub/FVgQkCj3g9pWfn0irfJDY0xzUMwE0c3YIPP4wxvm6/7gItFJAO3NvkNPvt6e1VYX4nIsT7XzKjlmgCIyBQRSRaR5MzMzHq8jJrtLiwhK7fI2jeMMc1GqBvHJwHTVDUROA14TURa4KY66eFVYd0MvCEibWu4zj688SZJqpqUkJAQ8MArrMt2kxva4D9jTHMRzNbcjUB3n8eJ3jZfV+G1UajqPBGJAuJVdRtQ5G1fKCJrgYO88xNruWaDSq2Y3NCmUzfGNBPBLHEsAPqLSG8RicA1fs+qdMx6YAyAiAwEooBMEUnwGtcRkT64RvBUVd0M7BKRI7zeVJcC/w3ia6hVxay4PTtY4jDGNA9BK3GoaqmIXA/MBsKAqaq6QkTuB5JVdRbwR+AFb7p2BS5XVRWR44D7RaQEKAeuUdUc79K/A6bhViH8mBAvY5uelUeXdlG0iggLZRjGGNNggjrwQFU/wjV6+267x+f+SuDoKs57F3i3mmsmA4cENtL9l5adZ4s3GWOalVA3jh/w0rLyrH3DGNOsWOKohx35xezIL7EeVcaYZsUSRz2kZdkcVcaY5scSRz2kZ9s648aY5scSRz2kZebRQqBHB5tuxBjTfFjiqIe07Hy6tW9FRLi9jcaY5sM+8eohPcu64hpjmh9LHPtJVW06dWNMs2SJYz9l5RaTW1RqPaqMMc2OJY79lG7LxRpjmilLHPupYgyHDf4zxjQ3ljj2U1pWHuEthMT2rUIdijHGNChLHPspPSuPHh1aEx5mb6ExpnmxT739lJaVZ+0bxphmyRLHfigvV9Zl59sYDmNMs2SJYz9s3V1IQUmZTadujGmWgpo4RGSciKwSkRQRuaOK/T1EZI6ILBKRpSJymrd9rIgsFJFl3s/RPufM9a652Lt1DOZrqIr1qDLGNGdBWwHQWzP8aWAskAEsEJFZ3qp/Fe4C3lbVZ0VkEG61wF5AFnCmqm4SkUNwy8928zlvsrcSYEikZ+UD0CveJjc0xjQ/wSxxjARSVDVVVYuB6cD4Ssco0Na73w7YBKCqi1R1k7d9BdBKRCKDGGudpGfnERHegq7trCuuMab5CWbi6AZs8Hmcwd6lBoD7gItFJANX2rihiuucC/ykqkU+2172qqnuFhGp6slFZIqIJItIcmZm5n6/iKqkZubRK641LVpU+dTGGNOkhbpxfBIwTVUTgdOA10RkT0wiMhh4BPitzzmTVXUIcKx3u6SqC6vq86qapKpJCQkJAQ06PdtmxTXGNF/BTBwbge4+jxO9bb6uAt4GUNV5QBQQDyAiicBM4FJVXVtxgqpu9H7uBt7AVYk1mLJyZX12vq36Z4xptoKZOBYA/UWkt4hEABOBWZWOWQ+MARCRgbjEkSkiscCHwB2q+l3FwSISLiIViaUlcAawPIivYR+bdhRQXFZuicMY02wFLXGoailwPa5H1M+43lMrROR+ETnLO+yPwNUisgR4E7hcVdU7rx9wT6Vut5HAbBFZCizGlWBeCNZrqEpFV1wbNW6Maa6C1h0XQFU/wjV6+267x+f+SuDoKs57AHigmsuOCGSMdVUxnbqVOIwxzVWoG8cPOGlZebSOCKNjTKPpHWyMMQ3KEkcdpXnrjFfTC9gYY5o8Sxx1lJ6VZ9VUxphmzRJHHZSUlbNhe4FNNWKMadYscdRBxvYCysqV3vHRoQ7FGGNCxhJHHaRl5QLQ20ocxphmzBJHHaRVzIpr040YY5oxSxx1kJ6VR0xUOB3aRIQ6FGOMCRlLHHWQnp1Hn3jrimuMad4scdRBamaeTTVijGn2LHH4qbCkjE07C6x9wxjT7Fni8NOGnHxUbY4qY4yxxOGn1Cyb3NAYY8ASh9/SbTp1Y4wBLHH4LT07jw5tImjXqmWoQzHGmJCyxOGnNJvc0BhjgCAnDhEZJyKrRCRFRO6oYn8PEZkjIotEZKmInOaz707vvFUicoq/1wyWiunUjTGmuQta4hCRMOBp4FRgEDBJRAZVOuwu3JKyh+LWJH/GO3eQ93gwMA54RkTC/LxmwOUXl7J1V5HNUWWMMQS3xDESSFHVVFUtBqYD4ysdo0Bb7347YJN3fzwwXVWLVDUNSPGu5881Ay69Yo4qq6oyxpigJo5uwAafxxneNl/3AReLSAZubfIbajnXn2sCICJTRCRZRJIzMzP39zUAts64Mcb4CnXj+CRgmqomAqcBr4lIQGJS1edVNUlVkxISEup1rbSKrrjWxmGMMYQH8dobge4+jxO9bb6uwrVhoKrzRCQKiK/l3NquGXBpWXl0jImkTWQw3y5jjDkwBLPEsQDoLyK9RSQC19g9q9Ix64ExACIyEIgCMr3jJopIpIj0BvoD8/28ZsClZ9nkhsYYUyFoX6FVtVRErgdmA2HAVFVdISL3A8mqOgv4I/CCiPwB11B+uaoqsEJE3gZWAqXAdapaBlDVNYP1GiqkZ+dx0sBOwX4aY4w5IAS17kVVP8I1evtuu8fn/krg6GrOfRB40J9rBtOuwhKycoutxGGMMZ5QN443eunWMG6MMXuxxFGLih5VfRIscRhjDFjiqFVaVh4i0KODjRo3xhiwxFGr9Kw8urZrRVTLsFCHYowxjYIljlqkZefTy+aoMsaYPSxx1EBVScvMtalGjDHGhyWOGmzPL2FXYan1qDLGGB+WOGqQZuuMG2PMPixx1MDWGTfGmH1Z4qhBenYeYS2E7u2tcdwYYypY4qhBalYeie1bERFub5MxxlSwecJrMKhLWxv4Z4wxlVjiqMF1J/YLdQjGGNPoWB2MMcaYOrHEYYwxpk4scRhjjKmToCYOERknIqtEJEVE7qhi/2Misti7rRaRHd72E322LxaRQhGZ4O2bJiJpPvuGB/M1GGOM2VvQGsdFJAx4GhgLZAALRGSWt+ofAKr6B5/jbwAO9bbPAYZ72zsAKcCnPpe/VVVnBCt2Y4wx1QtmiWMkkKKqqapaDEwHxtdw/CTgzSq2nwd8rKr5QYjRGGNMHQUzcXQDNvg8zvC27UNEegK9gS+r2D2RfRPKgyKy1KvqiqzmmlNEJFlEkjMzM+sevTHGmCo1lsbxicAMVS3z3SgiXYAhwGyfzXcCBwOHAx2A26u6oKo+r6pJqpqUkJAQnKiNMaYZCuYAwI1Ad5/Hid62qkwErqti+wXATFUtqdigqpu9u0Ui8jJwS22BLFy4MEtE1vkV9b7igaz9PLchWHz1Y/HVj8VXP409vp5VbQxm4lgA9BeR3riEMRG4qPJBInIw0B6YV8U1JuFKGL7Hd1HVzSIiwARgeW2BqOp+FzlEJFlVk/b3/GCz+OrH4qsfi69+Gnt81Qla4lDVUhG5HlfNFAZMVdUVInI/kKyqs7xDJwLTVVV9zxeRXrgSy1eVLv26iCQAAiwGrgnWazDGGLOvoM5VpaofAR9V2nZPpcf3VXNuOlU0pqvq6MBFaIwxpq4aS+N4Y/Z8qAOohcVXPxZf/Vh89dPY46uSVKohMsYYY2pkJQ5jjDF1YonDGGNMnVji8PgxIWOkiLzl7f/R6/XVULF1F5E5IrJSRFaIyE1VHHOCiOz0mfzxnqquFcQY00VkmffcyVXsFxF5wnv/lorIYQ0Y24BKk2buEpHfVzqmQd8/EZkqIttEZLnPtg4i8pmIrPF+tq/m3Mu8Y9aIyGUNGN8/ROQX7/c3U0Riqzm3xr+FIMZ3n4hs9PkdnlbNuTX+rwcxvrd8YksXkcXVnBv096/eVLXZ33DdhdcCfYAIYAkwqNIxvwOe8+5PBN5qwPi6AId592OA1VXEdwLwQQjfw3Qgvob9pwEf47pRHwH8GMLf9RagZyjfP+A44DBguc+2vwN3ePfvAB6p4rwOQKr3s713v30DxXcyEO7df6Sq+Pz5WwhifPcBt/jx+6/xfz1Y8VXa/0/gnlC9f/W9WYnD8WdCxvHAK979GcAYbxBi0KnqZlX9ybu/G/iZaub9asTGA6+q8wMQ600p09DGAGtVdX9nEggIVf0ayKm02fdv7BXcANfKTgE+U9UcVd0OfAaMa4j4VPVTVS31Hv6Amw0iJKp5//xR18lX90tN8XmfGxdQ9aSuBwRLHI4/EzLuOcb759kJxDVIdD68KrJDgR+r2H2kiCwRkY9FZHDDRoYCn4rIQhGZUsV+vye9DLKqJs2sEMr3D6CT/jqlzhagUxXHNJb38UpcCbIqtf0tBNP1XlXa1Gqq+hrD+3cssFVV11SzP5Tvn18scRxARCQaeBf4varuqrT7J1z1yzDgSeD9Bg7vGFU9DDgVuE5Ejmvg56+ViEQAZwHvVLE71O/fXtTVWTTKvvIi8megFHi9mkNC9bfwLNAXt5bPZlx1UGNU3RISFRr9/5IlDsefCRn3HCMi4UA7ILtBonPP2RKXNF5X1fcq71fVXaqa693/CGgpIvENFZ+qbvR+bgNm4qoEfNVl0stgORX4SVW3Vt4R6vfPs7Wi+s77ua2KY0L6PorI5cAZwGQvue3Dj7+FoFDVrapapqrlwAvVPG+o379w4BzgreqOCdX7VxeWOJw9EzJ630onArMqHTMLqOjBch7wZXX/OIHm1Ym+BPysqv+q5pjOFW0uIjIS97ttkMQmIm1EJKbiPq4RtfLkk7OAS73eVUcAO32qZRpKtd/0Qvn++fD9G7sM+G8Vx8wGThaR9l5VzMnsvexA0IjIOOA24CytZmE1P/8WghWfb5vZ2dU8rz//68F0EvCLqmZUtTOU71+dhLp1vrHccL1+VuN6XPzZ23Y/7p8EIApXxZECzAf6NGBsx+CqLZbiJnZc7MV7DXCNd8z1wApcL5EfgKMaML4+3vMu8WKoeP984xPcUsJrgWVAUgP/ftvgEkE7n20he/9wCWwzUIKrZ78K12b2BbAG+Bzo4B2bBLzoc+6V3t9hCnBFA8aXgmsfqPgbrOhl2BX4qKa/hQaK7zXvb2spLhl0qRyf93if//WGiM/bPq3ib87n2AZ//+p7sylHjDHG1IlVVRljjKkTSxzGGGPqxBKHMcaYOrHEYYwxpk4scRhjjKkTSxzGBICIlMneM/AGbNZVEenlO8uqMaEW1DXHjWlGClR1eKiDMKYhWInDmCDy1lb4u7e+wnwR6edt7yUiX3oT8n0hIj287Z28tS6WeLejvEuFicgL4tZj+VREWoXsRZlmzxKHMYHRqlJV1YU++3aq6hDgKeBxb9uTwCuqOhQ3WeAT3vYngK/UTbZ4GG70MEB/4GlVHQzsAM4N6qsxpgY2ctyYABCRXFWNrmJ7OjBaVVO9iSq3qGqciGThpsQo8bZvVtV4EckEElW1yOcavXBrcPT3Ht8OtFTVBxrgpRmzDytxGBN8Ws39uijyuV+GtU+aELLEYUzwXejzc553/3vczKwAk4FvvPtfANcCiEiYiLRrqCCN8Zd9azEmMFqJyGKfx5+oakWX3PYishRXapjkbbsBeFlEbgUygSu87TcBz4vIVbiSxbW4WVaNaTSsjcOYIPLaOJJUNSvUsRgTKFZVZYwxpk6sxGGMMaZOrMRhjDGmTixxGGOMqRNLHMYYY+rEEocxxpg6scRhjDGmTv4f2F8aZOBtmiEAAAAASUVORK5CYII=\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": [
    "import matplotlib.pyplot as plt\n",
    "plt.plot(history.history['accuracy'])\n",
    "plt.plot(history.history['val_accuracy'])\n",
    "plt.title('Model accuracy')\n",
    "plt.ylabel('Accuracy')\n",
    "plt.xlabel('Epoch')\n",
    "plt.legend(['Train', 'Test'], loc='upper left')\n",
    "plt.show()\n",
    "plt.plot(history.history['loss'])\n",
    "plt.plot(history.history['val_loss'])\n",
    "plt.title('Model loss')\n",
    "plt.ylabel('Loss')\n",
    "plt.xlabel('Epoch')\n",
    "plt.legend(['Train', 'Test'], loc='upper left')\n",
    "plt.show()"
   ]
  }
 ],
 "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": "3.7.10"
  },
  "papermill": {
   "default_parameters": {},
   "duration": 796.945679,
   "end_time": "2021-06-08T10:40:44.120131",
   "environment_variables": {},
   "exception": null,
   "input_path": "__notebook__.ipynb",
   "output_path": "__notebook__.ipynb",
   "parameters": {},
   "start_time": "2021-06-08T10:27:27.174452",
   "version": "2.3.3"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}