Challenge1-Grayscale.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "5654ed61-e81d-4eb7-941a-a8b5872e286e",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "2.6.0\n"
     ]
    }
   ],
   "source": [
    "import tensorflow as tf\n",
    "import numpy as np\n",
    "import os\n",
    "import random\n",
    "import pandas as pd\n",
    "import seaborn as sns\n",
    "import matplotlib as mpl\n",
    "import matplotlib.pyplot as plt\n",
    "from sklearn.model_selection import train_test_split\n",
    "from sklearn.metrics import accuracy_score, f1_score, precision_score, recall_score\n",
    "from sklearn.metrics import confusion_matrix\n",
    "from PIL import Image\n",
    "import splitfolders\n",
    "import visualkeras\n",
    "\n",
    "tfk = tf.keras\n",
    "tfkl = tf.keras.layers\n",
    "print(tf.__version__)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "26a3d493-6a07-4cce-a5d4-c2d3d8140678",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Random seed for reproducibility\n",
    "seed = 42\n",
    "\n",
    "random.seed(seed)\n",
    "os.environ['PYTHONHASHSEED'] = str(seed)\n",
    "np.random.seed(seed)\n",
    "tf.random.set_seed(seed)\n",
    "tf.compat.v1.set_random_seed(seed)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "0b55d4f0-bc65-43bd-af4f-6f23661e2220",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "/Users/irfanmasudi/ANNDL/Challenge\n"
     ]
    }
   ],
   "source": [
    "%cd /Users/irfanmasudi/ANNDL/Challenge/"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "997f5b41-482e-4730-a661-4484c19b3464",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Dataset folders \n",
    "dataset_dir = 'datasets'\n",
    "training_dir = os.path.join(dataset_dir, 'train')\n",
    "validation_dir = os.path.join(dataset_dir, 'val')\n",
    "#test_dir = os.path.join(dataset_dir, 'test')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "2da8af89-890a-49d4-a583-65fb21489b6b",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Plot example images from dataset\n",
    "labels = ['Apple',       # 0\n",
    "          'Blueberry',   # 1\n",
    "          'Cherry',      # 2\n",
    "          'Corn',        # 3\n",
    "          'Grape',       # 4\n",
    "          'Orange',      # 5\n",
    "          'Peach',       # 6\n",
    "          'Pepper',      # 7\n",
    "          'Potato',      # 8\n",
    "          'Raspberry',   # 9\n",
    "          'Soybean',     # 10\n",
    "          'Squash',      # 11\n",
    "          'Strawberry',  # 12\n",
    "          'Tomato']      # 13"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "144424cf-3d8a-4073-ae9b-0cdbfcc7e0c7",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Found 15062 images belonging to 14 classes.\n",
      "Found 2666 images belonging to 14 classes.\n"
     ]
    }
   ],
   "source": [
    "from tensorflow.keras.preprocessing.image import ImageDataGenerator\n",
    "\n",
    "valid_data_gen = ImageDataGenerator(rotation_range=40,\n",
    "                                        height_shift_range=60,\n",
    "                                        width_shift_range=60,\n",
    "                                        zoom_range=0.5,\n",
    "                                        horizontal_flip=True,\n",
    "                                        vertical_flip=True, \n",
    "                                        fill_mode='reflect',\n",
    "                                        rescale=1/255.)\n",
    "\n",
    "# Create an instance of ImageDataGenerator with Data Augmentation\n",
    "aug_train_data_gen = ImageDataGenerator(rotation_range=30,\n",
    "                                        height_shift_range=50,\n",
    "                                        width_shift_range=50,\n",
    "                                        zoom_range=0.3,\n",
    "                                        horizontal_flip=True,\n",
    "                                        vertical_flip=True, \n",
    "                                        fill_mode='reflect',\n",
    "                                        rescale=1/255.) # rescale value is multiplied to the image\n",
    "\n",
    "# Obtain a data generator with the 'ImageDataGenerator.flow_from_directory' method\n",
    "train_gen = aug_train_data_gen.flow_from_directory(directory=training_dir,\n",
    "                                                       target_size=(256,256),\n",
    "                                                       color_mode='grayscale',\n",
    "                                                       classes=None, # can be set to labels\n",
    "                                                       class_mode='categorical',\n",
    "                                                       batch_size=64,\n",
    "                                                       shuffle=True,\n",
    "                                                       seed=seed)\n",
    "\n",
    "valid_gen = valid_data_gen.flow_from_directory(directory=validation_dir,\n",
    "                                               target_size=(256,256),\n",
    "                                               color_mode='grayscale',\n",
    "                                               classes=None, # can be set to labels\n",
    "                                               class_mode='categorical',\n",
    "                                               batch_size=64,\n",
    "                                               shuffle=False,\n",
    "                                               seed=seed)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "9b711fb2-2796-4e58-95b0-61d7b2ef6607",
   "metadata": {},
   "outputs": [],
   "source": [
    "def get_next_batch(generator):\n",
    "    batch = next(generator)\n",
    "\n",
    "    image = batch[0]\n",
    "    target = batch[1]\n",
    "\n",
    "    print(\"(Input) image shape:\", image.shape)\n",
    "    print(\"Target shape:\",target.shape)\n",
    "\n",
    "    # Visualize only the first sample\n",
    "    image = image[0]\n",
    "    target = target[0]\n",
    "    target_idx = np.argmax(target)\n",
    "    print()\n",
    "    print(\"Categorical label:\", target)\n",
    "    print(\"Label:\", target_idx)\n",
    "    print(\"Class name:\", labels[target_idx])\n",
    "    \n",
    "    fig = plt.figure(figsize=(6, 4))\n",
    "    plt.imshow(np.uint8(image))\n",
    "    plt.show()\n",
    "    \n",
    "    return batch"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "2515b869-3088-4bf2-a3dd-8b4bca87ff30",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(Input) image shape: (64, 256, 256, 1)\n",
      "Target shape: (64, 14)\n",
      "\n",
      "Categorical label: [0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 1.]\n",
      "Label: 13\n",
      "Class name: Tomato\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 600x400 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Get a sample from dataset and show info\n",
    "_ = get_next_batch(train_gen)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "90784e78-52a8-43ab-988a-2b5c43a3f737",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{0: 1.282308871105057,\n",
       " 1: 2.716810966810967,\n",
       " 2: 2.1734487734487735,\n",
       " 3: 1.049616724738676,\n",
       " 4: 0.8683269918136747,\n",
       " 5: 0.7244829244829245,\n",
       " 6: 1.2962134251290878,\n",
       " 7: 1.6551648351648351,\n",
       " 8: 1.769501879699248,\n",
       " 9: 4.802933673469388,\n",
       " 10: 0.7835813130787639,\n",
       " 11: 2.2091522440598417,\n",
       " 12: 1.8808691308691308,\n",
       " 13: 0.22233046969562778}"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from sklearn.utils import compute_class_weight\n",
    "\n",
    "class_weights = compute_class_weight(class_weight = \"balanced\",\n",
    "                                     classes = np.unique(train_gen.classes),\n",
    "                                     y = train_gen.classes                                                    \n",
    "                                    )\n",
    "class_weights = dict(zip(np.unique(train_gen.classes), class_weights))\n",
    "class_weights"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "e650f6ca-76d4-4c01-b10a-11c9701f24c7",
   "metadata": {
    "tags": []
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Assigned labels\n",
      "{'Apple': 0, 'Blueberry': 1, 'Cherry': 2, 'Corn': 3, 'Grape': 4, 'Orange': 5, 'Peach': 6, 'Pepper': 7, 'Potato': 8, 'Raspberry': 9, 'Soybean': 10, 'Squash': 11, 'Strawberry': 12, 'Tomato': 13}\n",
      "\n",
      "Target classes\n",
      "[ 0  0  0 ... 13 13 13]\n"
     ]
    }
   ],
   "source": [
    "print(\"Assigned labels\")\n",
    "print(train_gen.class_indices)\n",
    "print()\n",
    "print(\"Target classes\")\n",
    "print(train_gen.classes)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "id": "eb45aebb-bb1d-4fe0-98c4-c64d7dc8ead4",
   "metadata": {
    "tags": []
   },
   "outputs": [],
   "source": [
    "input_shape = (256, 256, 1)\n",
    "epochs = 200"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "id": "893d1940-01dc-4826-8f70-f84294ed1ced",
   "metadata": {},
   "outputs": [],
   "source": [
    "METRICS = [\n",
    "      tfk.metrics.TruePositives(name='tp'),\n",
    "      tfk.metrics.FalsePositives(name='fp'),\n",
    "      tfk.metrics.TrueNegatives(name='tn'),\n",
    "      tfk.metrics.FalseNegatives(name='fn'), \n",
    "      tfk.metrics.BinaryAccuracy(name='accuracy'),\n",
    "      tfk.metrics.Precision(name='precision'),\n",
    "      tfk.metrics.Recall(name='recall'),\n",
    "      tfk.metrics.AUC(name='auc'),\n",
    "]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "id": "be8a85e8-f90f-48e8-9b5e-aff88147c6f5",
   "metadata": {
    "tags": []
   },
   "outputs": [],
   "source": [
    "# Model used for the exercise:\n",
    "# (Conv + ReLU + MaxPool) x 5 + FC x 2\n",
    "def build_model(input_shape):\n",
    "\n",
    "    # Build the neural network layer by layer\n",
    "    input_layer = tfkl.Input(shape=input_shape, name='Input')\n",
    "\n",
    "    conv1 = tfkl.Conv2D(\n",
    "        filters=16,\n",
    "        kernel_size=(3, 3),\n",
    "        strides = (1, 1),\n",
    "        padding = 'same',\n",
    "        kernel_initializer = tfk.initializers.GlorotUniform(seed)\n",
    "    )(input_layer)\n",
    "    \n",
    "    batchNorm1 = tfkl.BatchNormalization()(conv1)\n",
    "    \n",
    "    relu1 = tfk.activations.relu(batchNorm1)\n",
    "    \n",
    "    dropout1 = tfkl.Dropout(0.3)(relu1)\n",
    "    \n",
    "    pool1 = tfkl.MaxPooling2D(\n",
    "        pool_size = (2, 2)\n",
    "    )(dropout1)\n",
    "\n",
    "    conv2 = tfkl.Conv2D(\n",
    "        filters=32,\n",
    "        kernel_size=(3, 3),\n",
    "        strides = (1, 1),\n",
    "        padding = 'same',\n",
    "        kernel_initializer = tfk.initializers.GlorotUniform(seed)\n",
    "    )(pool1)\n",
    "\n",
    "    batchNorm2 = tfkl.BatchNormalization()(conv2)\n",
    "    \n",
    "    relu2 = tfk.activations.relu(batchNorm2)\n",
    "    \n",
    "    dropout2 = tfkl.Dropout(0.3)(relu2)\n",
    "    \n",
    "    pool2 = tfkl.MaxPooling2D(\n",
    "        pool_size = (2, 2)\n",
    "    )(dropout2)\n",
    "    \n",
    "    \n",
    "    conv3 = tfkl.Conv2D(\n",
    "        filters=64,\n",
    "        kernel_size=(3, 3),\n",
    "        strides = (1, 1),\n",
    "        padding = 'same',\n",
    "        kernel_initializer = tfk.initializers.GlorotUniform(seed)\n",
    "    )(pool2)\n",
    "\n",
    "    batchNorm3 = tfkl.BatchNormalization()(conv3)\n",
    "    \n",
    "    relu3 = tfk.activations.relu(batchNorm3)\n",
    "    \n",
    "    dropout3 = tfkl.Dropout(0.3)(relu3)\n",
    "    \n",
    "    pool3 = tfkl.MaxPooling2D(\n",
    "        pool_size = (2, 2)\n",
    "    )(dropout3)\n",
    "\n",
    "    conv4 = tfkl.Conv2D(\n",
    "        filters=128,\n",
    "        kernel_size=(3, 3),\n",
    "        strides = (1, 1),\n",
    "        padding = 'same',\n",
    "        activation = 'relu',\n",
    "        kernel_initializer = tfk.initializers.GlorotUniform(seed)\n",
    "    )(pool3)\n",
    "    \n",
    "    batchNorm4 = tfkl.BatchNormalization()(conv4)\n",
    "    \n",
    "    relu4 = tfk.activations.relu(batchNorm4)\n",
    "    \n",
    "    dropout4 = tfkl.Dropout(0.3)(relu4)\n",
    "    \n",
    "    pool4 = tfkl.MaxPooling2D(\n",
    "        pool_size = (2, 2)\n",
    "    )(dropout4)\n",
    "\n",
    "    conv5 = tfkl.Conv2D(\n",
    "        filters=256,\n",
    "        kernel_size=(3, 3),\n",
    "        strides = (1, 1),\n",
    "        padding = 'same',\n",
    "        kernel_initializer = tfk.initializers.GlorotUniform(seed)\n",
    "    )(pool4)\n",
    "    \n",
    "    batchNorm5 = tfkl.BatchNormalization()(conv5)\n",
    "    \n",
    "    relu5 = tfk.activations.relu(batchNorm5)\n",
    "    \n",
    "    glob_pooling = tfkl.GlobalAveragePooling2D(name='GlobalAvglPooling')(relu5)\n",
    "    glob_pooling = tfkl.Dropout(0.3, seed=seed, name='GAPDropout')(glob_pooling)\n",
    "    \n",
    "    dense1 = tfkl.Dense(\n",
    "        units=256,  \n",
    "        activation='relu',\n",
    "        kernel_initializer = tfk.initializers.GlorotUniform(seed),\n",
    "        name='dense1')(glob_pooling)\n",
    "    dense1 = tfkl.Dropout(0.3, seed=seed, name='DenseDropout')(dense1)\n",
    "\n",
    "    classifier_layer = tfkl.Dense(\n",
    "        units=256,  \n",
    "        activation='relu',\n",
    "        kernel_initializer = tfk.initializers.GlorotUniform(seed),\n",
    "        name='Classifier')(dense1)\n",
    "    classifier_layer = tfkl.Dropout(0.3, seed=seed, name='ClassifierDropout')(classifier_layer)\n",
    "\n",
    "    output_layer = tfkl.Dense(\n",
    "        units=14, \n",
    "        activation='softmax', \n",
    "        kernel_initializer = tfk.initializers.GlorotUniform(seed),\n",
    "        name='Output')(classifier_layer)\n",
    "\n",
    "    # Connect input and output through the Model class\n",
    "    model = tfk.Model(inputs=input_layer, outputs=output_layer, name='model_GAP_BN')\n",
    "\n",
    "    # Compile the model\n",
    "    model.compile(loss=tfk.losses.CategoricalCrossentropy(), optimizer=tfk.optimizers.Adam(), metrics=METRICS)\n",
    "\n",
    "    # Return the model\n",
    "    return model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "id": "9f7f3aa1-3fdb-483f-a041-f8d10dced1c3",
   "metadata": {
    "tags": []
   },
   "outputs": [],
   "source": [
    "# Utility function to create folders and callbacks for training\n",
    "from datetime import datetime\n",
    "\n",
    "def create_folders_and_callbacks(model_name):\n",
    "    exps_dir = os.path.join('logs/Grayscale')\n",
    "    if not os.path.exists(exps_dir):\n",
    "        os.makedirs(exps_dir)\n",
    "\n",
    "    now = datetime.now().strftime('%b%d_%H-%M-%S')\n",
    "\n",
    "    exp_dir = os.path.join(exps_dir, model_name + '_' + str(now))\n",
    "    if not os.path.exists(exp_dir):\n",
    "        os.makedirs(exp_dir)\n",
    "      \n",
    "    callbacks = []\n",
    "\n",
    "  # Model checkpoint\n",
    "  # ----------------\n",
    "    ckpt_dir = os.path.join(exp_dir, 'ckpts')\n",
    "    if not os.path.exists(ckpt_dir):\n",
    "        os.makedirs(ckpt_dir)\n",
    "\n",
    "    ckpt_callback = tf.keras.callbacks.ModelCheckpoint(filepath=os.path.join(ckpt_dir, 'cp.ckpt'), \n",
    "                                                     save_weights_only=False, # True to save only weights\n",
    "                                                     save_best_only=False) # True to save only the best epoch \n",
    "    callbacks.append(ckpt_callback)\n",
    "\n",
    "  # Visualize Learning on Tensorboard\n",
    "  # ---------------------------------\n",
    "    tb_dir = os.path.join(exp_dir, 'tb_logs')\n",
    "    if not os.path.exists(tb_dir):\n",
    "        os.makedirs(tb_dir)\n",
    "      \n",
    "  # By default shows losses and metrics for both training and validation\n",
    "    tb_callback = tf.keras.callbacks.TensorBoard(log_dir=tb_dir, \n",
    "                                               profile_batch=0,\n",
    "                                               histogram_freq=1)  # if > 0 (epochs) shows weights histograms\n",
    "    callbacks.append(tb_callback)\n",
    "\n",
    "  # Early Stopping\n",
    "  # --------------\n",
    "    es_callback = tf.keras.callbacks.EarlyStopping(monitor='val_auc', patience=10, \n",
    "                                                   restore_best_weights=True, \n",
    "                                                   verbose=1,\n",
    "                                                   mode='max',)\n",
    "    callbacks.append(es_callback)\n",
    "\n",
    "    return callbacks"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "id": "848fccf3-14f6-424c-84a9-e6267eff870d",
   "metadata": {
    "tags": []
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model: \"model_GAP_BN\"\n",
      "_________________________________________________________________\n",
      "Layer (type)                 Output Shape              Param #   \n",
      "=================================================================\n",
      "Input (InputLayer)           [(None, 256, 256, 1)]     0         \n",
      "_________________________________________________________________\n",
      "conv2d (Conv2D)              (None, 256, 256, 16)      160       \n",
      "_________________________________________________________________\n",
      "batch_normalization (BatchNo (None, 256, 256, 16)      64        \n",
      "_________________________________________________________________\n",
      "tf.nn.relu (TFOpLambda)      (None, 256, 256, 16)      0         \n",
      "_________________________________________________________________\n",
      "dropout (Dropout)            (None, 256, 256, 16)      0         \n",
      "_________________________________________________________________\n",
      "max_pooling2d (MaxPooling2D) (None, 128, 128, 16)      0         \n",
      "_________________________________________________________________\n",
      "conv2d_1 (Conv2D)            (None, 128, 128, 32)      4640      \n",
      "_________________________________________________________________\n",
      "batch_normalization_1 (Batch (None, 128, 128, 32)      128       \n",
      "_________________________________________________________________\n",
      "tf.nn.relu_1 (TFOpLambda)    (None, 128, 128, 32)      0         \n",
      "_________________________________________________________________\n",
      "dropout_1 (Dropout)          (None, 128, 128, 32)      0         \n",
      "_________________________________________________________________\n",
      "max_pooling2d_1 (MaxPooling2 (None, 64, 64, 32)        0         \n",
      "_________________________________________________________________\n",
      "conv2d_2 (Conv2D)            (None, 64, 64, 64)        18496     \n",
      "_________________________________________________________________\n",
      "batch_normalization_2 (Batch (None, 64, 64, 64)        256       \n",
      "_________________________________________________________________\n",
      "tf.nn.relu_2 (TFOpLambda)    (None, 64, 64, 64)        0         \n",
      "_________________________________________________________________\n",
      "dropout_2 (Dropout)          (None, 64, 64, 64)        0         \n",
      "_________________________________________________________________\n",
      "max_pooling2d_2 (MaxPooling2 (None, 32, 32, 64)        0         \n",
      "_________________________________________________________________\n",
      "conv2d_3 (Conv2D)            (None, 32, 32, 128)       73856     \n",
      "_________________________________________________________________\n",
      "batch_normalization_3 (Batch (None, 32, 32, 128)       512       \n",
      "_________________________________________________________________\n",
      "tf.nn.relu_3 (TFOpLambda)    (None, 32, 32, 128)       0         \n",
      "_________________________________________________________________\n",
      "dropout_3 (Dropout)          (None, 32, 32, 128)       0         \n",
      "_________________________________________________________________\n",
      "max_pooling2d_3 (MaxPooling2 (None, 16, 16, 128)       0         \n",
      "_________________________________________________________________\n",
      "conv2d_4 (Conv2D)            (None, 16, 16, 256)       295168    \n",
      "_________________________________________________________________\n",
      "batch_normalization_4 (Batch (None, 16, 16, 256)       1024      \n",
      "_________________________________________________________________\n",
      "tf.nn.relu_4 (TFOpLambda)    (None, 16, 16, 256)       0         \n",
      "_________________________________________________________________\n",
      "GlobalAvglPooling (GlobalAve (None, 256)               0         \n",
      "_________________________________________________________________\n",
      "GAPDropout (Dropout)         (None, 256)               0         \n",
      "_________________________________________________________________\n",
      "dense1 (Dense)               (None, 256)               65792     \n",
      "_________________________________________________________________\n",
      "DenseDropout (Dropout)       (None, 256)               0         \n",
      "_________________________________________________________________\n",
      "Classifier (Dense)           (None, 256)               65792     \n",
      "_________________________________________________________________\n",
      "ClassifierDropout (Dropout)  (None, 256)               0         \n",
      "_________________________________________________________________\n",
      "Output (Dense)               (None, 14)                3598      \n",
      "=================================================================\n",
      "Total params: 529,486\n",
      "Trainable params: 528,494\n",
      "Non-trainable params: 992\n",
      "_________________________________________________________________\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<PIL.Image.Image image mode=RGBA size=1037x378 at 0x1638766D0>"
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "model = build_model(input_shape)\n",
    "model.summary()\n",
    "visualkeras.layered_view(model, legend=True, spacing=10, scale_xy=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "id": "abd37bea-923f-4416-9f79-86d7b5fec523",
   "metadata": {
    "tags": []
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/200\n",
      "236/236 [==============================] - 130s 544ms/step - loss: 1.9068 - tp: 1643.0000 - fp: 1216.0000 - tn: 194590.0000 - fn: 13419.0000 - accuracy: 0.9306 - precision: 0.5747 - recall: 0.1091 - auc: 0.8255 - val_loss: 4.3395 - val_tp: 3.0000 - val_fp: 24.0000 - val_tn: 34634.0000 - val_fn: 2663.0000 - val_accuracy: 0.9280 - val_precision: 0.1111 - val_recall: 0.0011 - val_auc: 0.5478\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "WARNING:absl:Function `_wrapped_model` contains input name(s) Input with unsupported characters which will be renamed to input in the SavedModel.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 2/200\n",
      "236/236 [==============================] - 132s 557ms/step - loss: 1.4408 - tp: 4199.0000 - fp: 2156.0000 - tn: 193650.0000 - fn: 10863.0000 - accuracy: 0.9383 - precision: 0.6607 - recall: 0.2788 - auc: 0.9014 - val_loss: 4.8302 - val_tp: 155.0000 - val_fp: 1196.0000 - val_tn: 33462.0000 - val_fn: 2511.0000 - val_accuracy: 0.9007 - val_precision: 0.1147 - val_recall: 0.0581 - val_auc: 0.5608\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "WARNING:absl:Function `_wrapped_model` contains input name(s) Input with unsupported characters which will be renamed to input in the SavedModel.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 3/200\n",
      "236/236 [==============================] - 138s 586ms/step - loss: 1.1541 - tp: 6109.0000 - fp: 2507.0000 - tn: 193299.0000 - fn: 8953.0000 - accuracy: 0.9457 - precision: 0.7090 - recall: 0.4056 - auc: 0.9355 - val_loss: 3.0704 - val_tp: 672.0000 - val_fp: 687.0000 - val_tn: 33971.0000 - val_fn: 1994.0000 - val_accuracy: 0.9282 - val_precision: 0.4945 - val_recall: 0.2521 - val_auc: 0.7232\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "WARNING:absl:Function `_wrapped_model` contains input name(s) Input with unsupported characters which will be renamed to input in the SavedModel.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 4/200\n",
      "236/236 [==============================] - 144s 608ms/step - loss: 0.9313 - tp: 8023.0000 - fp: 2614.0000 - tn: 193192.0000 - fn: 7039.0000 - accuracy: 0.9542 - precision: 0.7543 - recall: 0.5327 - auc: 0.9561 - val_loss: 2.8713 - val_tp: 500.0000 - val_fp: 1087.0000 - val_tn: 33571.0000 - val_fn: 2166.0000 - val_accuracy: 0.9128 - val_precision: 0.3151 - val_recall: 0.1875 - val_auc: 0.7441\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "WARNING:absl:Function `_wrapped_model` contains input name(s) Input with unsupported characters which will be renamed to input in the SavedModel.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 5/200\n",
      "236/236 [==============================] - 131s 556ms/step - loss: 0.8130 - tp: 8932.0000 - fp: 2551.0000 - tn: 193255.0000 - fn: 6130.0000 - accuracy: 0.9588 - precision: 0.7778 - recall: 0.5930 - auc: 0.9642 - val_loss: 4.4622 - val_tp: 444.0000 - val_fp: 1985.0000 - val_tn: 32673.0000 - val_fn: 2222.0000 - val_accuracy: 0.8873 - val_precision: 0.1828 - val_recall: 0.1665 - val_auc: 0.6540\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "WARNING:absl:Function `_wrapped_model` contains input name(s) Input with unsupported characters which will be renamed to input in the SavedModel.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 6/200\n",
      "236/236 [==============================] - 136s 575ms/step - loss: 0.7166 - tp: 9650.0000 - fp: 2361.0000 - tn: 193445.0000 - fn: 5412.0000 - accuracy: 0.9631 - precision: 0.8034 - recall: 0.6407 - auc: 0.9704 - val_loss: 2.9307 - val_tp: 628.0000 - val_fp: 1557.0000 - val_tn: 33101.0000 - val_fn: 2038.0000 - val_accuracy: 0.9037 - val_precision: 0.2874 - val_recall: 0.2356 - val_auc: 0.7778\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "WARNING:absl:Function `_wrapped_model` contains input name(s) Input with unsupported characters which will be renamed to input in the SavedModel.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 7/200\n",
      "236/236 [==============================] - 129s 545ms/step - loss: 0.6600 - tp: 10153.0000 - fp: 2317.0000 - tn: 193489.0000 - fn: 4909.0000 - accuracy: 0.9657 - precision: 0.8142 - recall: 0.6741 - auc: 0.9739 - val_loss: 3.8270 - val_tp: 434.0000 - val_fp: 1784.0000 - val_tn: 32874.0000 - val_fn: 2232.0000 - val_accuracy: 0.8924 - val_precision: 0.1957 - val_recall: 0.1628 - val_auc: 0.7085\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "WARNING:absl:Function `_wrapped_model` contains input name(s) Input with unsupported characters which will be renamed to input in the SavedModel.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 8/200\n",
      "236/236 [==============================] - 136s 574ms/step - loss: 0.5716 - tp: 10817.0000 - fp: 2139.0000 - tn: 193667.0000 - fn: 4245.0000 - accuracy: 0.9697 - precision: 0.8349 - recall: 0.7182 - auc: 0.9791 - val_loss: 1.9662 - val_tp: 891.0000 - val_fp: 900.0000 - val_tn: 33758.0000 - val_fn: 1775.0000 - val_accuracy: 0.9283 - val_precision: 0.4975 - val_recall: 0.3342 - val_auc: 0.8573\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "WARNING:absl:Function `_wrapped_model` contains input name(s) Input with unsupported characters which will be renamed to input in the SavedModel.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 9/200\n",
      "236/236 [==============================] - 128s 540ms/step - loss: 0.5419 - tp: 11064.0000 - fp: 2053.0000 - tn: 193753.0000 - fn: 3998.0000 - accuracy: 0.9713 - precision: 0.8435 - recall: 0.7346 - auc: 0.9810 - val_loss: 3.8579 - val_tp: 510.0000 - val_fp: 1834.0000 - val_tn: 32824.0000 - val_fn: 2156.0000 - val_accuracy: 0.8931 - val_precision: 0.2176 - val_recall: 0.1913 - val_auc: 0.7292\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "WARNING:absl:Function `_wrapped_model` contains input name(s) Input with unsupported characters which will be renamed to input in the SavedModel.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 10/200\n",
      "236/236 [==============================] - 126s 532ms/step - loss: 0.5028 - tp: 11285.0000 - fp: 1989.0000 - tn: 193817.0000 - fn: 3777.0000 - accuracy: 0.9727 - precision: 0.8502 - recall: 0.7492 - auc: 0.9827 - val_loss: 4.1782 - val_tp: 531.0000 - val_fp: 1900.0000 - val_tn: 32758.0000 - val_fn: 2135.0000 - val_accuracy: 0.8919 - val_precision: 0.2184 - val_recall: 0.1992 - val_auc: 0.7237\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "WARNING:absl:Function `_wrapped_model` contains input name(s) Input with unsupported characters which will be renamed to input in the SavedModel.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 11/200\n",
      "236/236 [==============================] - 129s 544ms/step - loss: 0.4764 - tp: 11540.0000 - fp: 1945.0000 - tn: 193861.0000 - fn: 3522.0000 - accuracy: 0.9741 - precision: 0.8558 - recall: 0.7662 - auc: 0.9843 - val_loss: 3.6150 - val_tp: 589.0000 - val_fp: 1694.0000 - val_tn: 32964.0000 - val_fn: 2077.0000 - val_accuracy: 0.8990 - val_precision: 0.2580 - val_recall: 0.2209 - val_auc: 0.7335\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "WARNING:absl:Function `_wrapped_model` contains input name(s) Input with unsupported characters which will be renamed to input in the SavedModel.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 12/200\n",
      "236/236 [==============================] - 125s 529ms/step - loss: 0.4410 - tp: 11773.0000 - fp: 1824.0000 - tn: 193982.0000 - fn: 3289.0000 - accuracy: 0.9758 - precision: 0.8659 - recall: 0.7816 - auc: 0.9859 - val_loss: 5.6195 - val_tp: 394.0000 - val_fp: 2149.0000 - val_tn: 32509.0000 - val_fn: 2272.0000 - val_accuracy: 0.8816 - val_precision: 0.1549 - val_recall: 0.1478 - val_auc: 0.6359\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "WARNING:absl:Function `_wrapped_model` contains input name(s) Input with unsupported characters which will be renamed to input in the SavedModel.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 13/200\n",
      "236/236 [==============================] - 140s 594ms/step - loss: 0.4186 - tp: 11991.0000 - fp: 1734.0000 - tn: 194072.0000 - fn: 3071.0000 - accuracy: 0.9772 - precision: 0.8737 - recall: 0.7961 - auc: 0.9872 - val_loss: 5.5941 - val_tp: 345.0000 - val_fp: 2181.0000 - val_tn: 32477.0000 - val_fn: 2321.0000 - val_accuracy: 0.8794 - val_precision: 0.1366 - val_recall: 0.1294 - val_auc: 0.6435\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "WARNING:absl:Function `_wrapped_model` contains input name(s) Input with unsupported characters which will be renamed to input in the SavedModel.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 14/200\n",
      "236/236 [==============================] - 129s 545ms/step - loss: 0.3826 - tp: 12132.0000 - fp: 1656.0000 - tn: 194150.0000 - fn: 2930.0000 - accuracy: 0.9783 - precision: 0.8799 - recall: 0.8055 - auc: 0.9878 - val_loss: 4.3193 - val_tp: 476.0000 - val_fp: 1990.0000 - val_tn: 32668.0000 - val_fn: 2190.0000 - val_accuracy: 0.8880 - val_precision: 0.1930 - val_recall: 0.1785 - val_auc: 0.7195\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "WARNING:absl:Function `_wrapped_model` contains input name(s) Input with unsupported characters which will be renamed to input in the SavedModel.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 15/200\n",
      "236/236 [==============================] - 134s 567ms/step - loss: 0.3580 - tp: 12453.0000 - fp: 1568.0000 - tn: 194238.0000 - fn: 2609.0000 - accuracy: 0.9802 - precision: 0.8882 - recall: 0.8268 - auc: 0.9896 - val_loss: 8.2337 - val_tp: 319.0000 - val_fp: 2325.0000 - val_tn: 32333.0000 - val_fn: 2347.0000 - val_accuracy: 0.8748 - val_precision: 0.1207 - val_recall: 0.1197 - val_auc: 0.5815\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "WARNING:absl:Function `_wrapped_model` contains input name(s) Input with unsupported characters which will be renamed to input in the SavedModel.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 16/200\n",
      "117/236 [=============>................] - ETA: 1:02 - loss: 0.3423 - tp: 6121.0000 - fp: 778.0000 - tn: 96020.0000 - fn: 1325.0000 - accuracy: 0.9798 - precision: 0.8872 - recall: 0.8221 - auc: 0.9899"
     ]
    },
    {
     "ename": "KeyboardInterrupt",
     "evalue": "",
     "output_type": "error",
     "traceback": [
      "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
      "\u001b[0;31mKeyboardInterrupt\u001b[0m                         Traceback (most recent call last)",
      "\u001b[0;32m<ipython-input-16-d98f331845c3>\u001b[0m in \u001b[0;36m<module>\u001b[0;34m\u001b[0m\n\u001b[1;32m      6\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m      7\u001b[0m \u001b[0;31m# Train the model\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m----> 8\u001b[0;31m history = model.fit(\n\u001b[0m\u001b[1;32m      9\u001b[0m     \u001b[0mx\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mtrain_gen\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     10\u001b[0m     \u001b[0mepochs\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mepochs\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m~/miniforge3/envs/tensorflow/lib/python3.9/site-packages/keras/engine/training.py\u001b[0m in \u001b[0;36mfit\u001b[0;34m(self, x, y, batch_size, epochs, verbose, callbacks, validation_split, validation_data, shuffle, class_weight, sample_weight, initial_epoch, steps_per_epoch, validation_steps, validation_batch_size, validation_freq, max_queue_size, workers, use_multiprocessing)\u001b[0m\n\u001b[1;32m   1182\u001b[0m                 _r=1):\n\u001b[1;32m   1183\u001b[0m               \u001b[0mcallbacks\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mon_train_batch_begin\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mstep\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m-> 1184\u001b[0;31m               \u001b[0mtmp_logs\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mtrain_function\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0miterator\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m   1185\u001b[0m               \u001b[0;32mif\u001b[0m \u001b[0mdata_handler\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mshould_sync\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   1186\u001b[0m                 \u001b[0mcontext\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0masync_wait\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m~/miniforge3/envs/tensorflow/lib/python3.9/site-packages/tensorflow/python/eager/def_function.py\u001b[0m in \u001b[0;36m__call__\u001b[0;34m(self, *args, **kwds)\u001b[0m\n\u001b[1;32m    883\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    884\u001b[0m       \u001b[0;32mwith\u001b[0m \u001b[0mOptionalXlaContext\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_jit_compile\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 885\u001b[0;31m         \u001b[0mresult\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_call\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m*\u001b[0m\u001b[0margs\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m**\u001b[0m\u001b[0mkwds\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    886\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    887\u001b[0m       \u001b[0mnew_tracing_count\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mexperimental_get_tracing_count\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m~/miniforge3/envs/tensorflow/lib/python3.9/site-packages/tensorflow/python/eager/def_function.py\u001b[0m in \u001b[0;36m_call\u001b[0;34m(self, *args, **kwds)\u001b[0m\n\u001b[1;32m    915\u001b[0m       \u001b[0;31m# In this case we have created variables on the first call, so we run the\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    916\u001b[0m       \u001b[0;31m# defunned version which is guaranteed to never create variables.\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 917\u001b[0;31m       \u001b[0;32mreturn\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_stateless_fn\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m*\u001b[0m\u001b[0margs\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m**\u001b[0m\u001b[0mkwds\u001b[0m\u001b[0;34m)\u001b[0m  \u001b[0;31m# pylint: disable=not-callable\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    918\u001b[0m     \u001b[0;32melif\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_stateful_fn\u001b[0m \u001b[0;32mis\u001b[0m \u001b[0;32mnot\u001b[0m \u001b[0;32mNone\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    919\u001b[0m       \u001b[0;31m# Release the lock early so that multiple threads can perform the call\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m~/miniforge3/envs/tensorflow/lib/python3.9/site-packages/tensorflow/python/eager/function.py\u001b[0m in \u001b[0;36m__call__\u001b[0;34m(self, *args, **kwargs)\u001b[0m\n\u001b[1;32m   3037\u001b[0m       (graph_function,\n\u001b[1;32m   3038\u001b[0m        filtered_flat_args) = self._maybe_define_function(args, kwargs)\n\u001b[0;32m-> 3039\u001b[0;31m     return graph_function._call_flat(\n\u001b[0m\u001b[1;32m   3040\u001b[0m         filtered_flat_args, captured_inputs=graph_function.captured_inputs)  # pylint: disable=protected-access\n\u001b[1;32m   3041\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m~/miniforge3/envs/tensorflow/lib/python3.9/site-packages/tensorflow/python/eager/function.py\u001b[0m in \u001b[0;36m_call_flat\u001b[0;34m(self, args, captured_inputs, cancellation_manager)\u001b[0m\n\u001b[1;32m   1961\u001b[0m         and executing_eagerly):\n\u001b[1;32m   1962\u001b[0m       \u001b[0;31m# No tape is watching; skip to running the function.\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m-> 1963\u001b[0;31m       return self._build_call_outputs(self._inference_function.call(\n\u001b[0m\u001b[1;32m   1964\u001b[0m           ctx, args, cancellation_manager=cancellation_manager))\n\u001b[1;32m   1965\u001b[0m     forward_backward = self._select_forward_and_backward_functions(\n",
      "\u001b[0;32m~/miniforge3/envs/tensorflow/lib/python3.9/site-packages/tensorflow/python/eager/function.py\u001b[0m in \u001b[0;36mcall\u001b[0;34m(self, ctx, args, cancellation_manager)\u001b[0m\n\u001b[1;32m    589\u001b[0m       \u001b[0;32mwith\u001b[0m \u001b[0m_InterpolateFunctionError\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    590\u001b[0m         \u001b[0;32mif\u001b[0m \u001b[0mcancellation_manager\u001b[0m \u001b[0;32mis\u001b[0m \u001b[0;32mNone\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 591\u001b[0;31m           outputs = execute.execute(\n\u001b[0m\u001b[1;32m    592\u001b[0m               \u001b[0mstr\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0msignature\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mname\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    593\u001b[0m               \u001b[0mnum_outputs\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_num_outputs\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m~/miniforge3/envs/tensorflow/lib/python3.9/site-packages/tensorflow/python/eager/execute.py\u001b[0m in \u001b[0;36mquick_execute\u001b[0;34m(op_name, num_outputs, inputs, attrs, ctx, name)\u001b[0m\n\u001b[1;32m     57\u001b[0m   \u001b[0;32mtry\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     58\u001b[0m     \u001b[0mctx\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mensure_initialized\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m---> 59\u001b[0;31m     tensors = pywrap_tfe.TFE_Py_Execute(ctx._handle, device_name, op_name,\n\u001b[0m\u001b[1;32m     60\u001b[0m                                         inputs, attrs, num_outputs)\n\u001b[1;32m     61\u001b[0m   \u001b[0;32mexcept\u001b[0m \u001b[0mcore\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_NotOkStatusException\u001b[0m \u001b[0;32mas\u001b[0m \u001b[0me\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;31mKeyboardInterrupt\u001b[0m: "
     ]
    }
   ],
   "source": [
    "tf.get_logger().setLevel('WARNING') #  if you want to suppress only INFOs\n",
    "tf.get_logger().setLevel('ERROR') #  if you want to suppress both WARNINGs and INFOs\n",
    "\n",
    "# Create folders and callbacks and fit\n",
    "train_callbacks = create_folders_and_callbacks(model_name='BatchNormGray')\n",
    "\n",
    "# Train the model\n",
    "history = model.fit(\n",
    "    x = train_gen,\n",
    "    epochs = epochs,\n",
    "    validation_data = valid_gen,\n",
    "    callbacks = train_callbacks,\n",
    "    class_weight = class_weights,\n",
    ").history"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "ed9c10d8-d94e-40b5-8044-83377fb7268c",
   "metadata": {
    "tags": []
   },
   "outputs": [],
   "source": [
    "# Save best epoch model\n",
    "model.save(\"best_model/GAPLossData\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "id": "a0dbfa0e-f65f-4075-b516-c31d47c7ea49",
   "metadata": {},
   "outputs": [],
   "source": [
    "del model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "id": "cc05b392-8d74-47bb-b73b-b012bc5047af",
   "metadata": {},
   "outputs": [],
   "source": [
    "model = tfk.models.load_model(\"logs/GAPBatchNorm/Batch_Norm_Nov19_12-55-38/ckpts/cp.ckpt\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "303df48e-349c-4a17-a63f-99c03fba9d0f",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1779/1779 [==============================] - 14s 8ms/step - loss: 0.1587 - accuracy: 0.9545\n",
      "\n",
      "Test metrics with data augmentation\n",
      "{'loss': 0.15866169333457947, 'accuracy': 0.9544687867164612}\n"
     ]
    }
   ],
   "source": [
    "model_test_metrics = model.evaluate(valid_gen, return_dict=True)\n",
    "\n",
    "print()\n",
    "print(\"Test metrics with data augmentation\")\n",
    "print(model_test_metrics)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "id": "f043e550-bd3b-4ed6-81f5-ffb6b7bcf55b",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Confusion Matrix\n",
      "[[113   0  12   2   0   0   4   6   0   0  11   0   0   1]\n",
      " [  0  65   4   0   0   0   2   0   0   0   0   0   0   0]\n",
      " [  1   2  71   0   0   0   3   3   0   0   8   0   0   0]\n",
      " [  0   0   4 172   0   0   0   2   0   0   3   0   0   0]\n",
      " [  6   0   3   1 159   0   4   4   0   1   3   0  22  16]\n",
      " [  3   3  25   0   0 152   5  19   2   0  47   0   0   7]\n",
      " [  0   0   2   0   0   0 141   0   0   0   2   0   0   2]\n",
      " [  0   0   4   0   1   0   2 102   0   0   6   0   0   0]\n",
      " [  2   0   1   1   1   0   1   0  66   0  11   0   1  24]\n",
      " [  0   0   0   0   0   0   0   0   0  37   1   0   1   1]\n",
      " [  1   0   0   0   0   0   0   1   0   1 240   0   0   0]\n",
      " [  2   1  14   6   0   0   0   0   0   0   1  61   0   2]\n",
      " [  0   1   1   3   3   0   1   0   0   2   2   0  86   2]\n",
      " [ 11   5  23   3   4   1  24  14   9   8  37   0   1 714]]\n"
     ]
    }
   ],
   "source": [
    "predictions = model.predict(valid_gen)\n",
    "y_pred = np.argmax(predictions, axis=1)\n",
    "print('Confusion Matrix')\n",
    "print(confusion_matrix(valid_gen.classes, y_pred))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "id": "e287f3a7-5d53-485c-9cd1-958452303333",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Confusion Matrix\n",
      "[[109   0   0   0   9   0   4   9   3   1   4   0   0  10]\n",
      " [  0  57   1   0   2   0   0   8   2   0   0   0   1   0]\n",
      " [  0   0  60   0   7   1   1  10   0   0   4   0   0   5]\n",
      " [  0   0   0 172   0   0   0   0   0   0   0   0   9   0]\n",
      " [  0   0   0   0 202   0   0   2   6   0   1   0   7   1]\n",
      " [  0   0   0   5   0 229   3   8  11   0   1   0   0   6]\n",
      " [  0   0   2   1   1   1 135   5   0   0   0   0   0   2]\n",
      " [  0   0   0   0   1   2   0 110   1   0   0   1   0   0]\n",
      " [  0   0   0   0   0   0   0   0  96   0   0   0   7   5]\n",
      " [  0   0   0   0   1   0   0   0   2  34   0   0   2   1]\n",
      " [  1   1   0   1   0   0   0   1  14   0 211   0   0  14]\n",
      " [  0   0   0   5   0   0   0   1   1   0   0  78   0   2]\n",
      " [  0   0   0   0   9   0   0   5   4   2   1   0  79   1]\n",
      " [  2   0   0   3  24   4   0   3  56   1   1   0   6 754]]\n"
     ]
    }
   ],
   "source": [
    "predictions = model.predict(valid_gen)\n",
    "y_pred = np.argmax(predictions, axis=1)\n",
    "print('Confusion Matrix')\n",
    "print(confusion_matrix(valid_gen.classes, y_pred))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "id": "55e4128e-6213-4a6a-973e-343d21734447",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Accuracy: 0.8725\n",
      "Precision: 0.8698\n",
      "Recall: 0.8574\n",
      "F1: 0.8532\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAsgAAAJMCAYAAADnilC4AAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjQuMiwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy8rg+JYAAAACXBIWXMAAA9hAAAPYQGoP6dpAACyKUlEQVR4nOzdd1hU19YG8HdoQwdFBVGkBFQQVJSIqLFEEU0saK5YoygaEyv2EAtggWhiiRq7giW2WLiWxB6NSoxIxErsigXEgqKIM8Cc7w8/z2UcMGCYgvP+8pzngT179lkzSlysWWcfiSAIAoiIiIiICABgoO0AiIiIiIh0CRNkIiIiIqJCmCATERERERXCBJmIiIiIqBAmyEREREREhTBBJiIiIiIqhAkyEREREVEhTJCJiIiIiAphgkxEREREVIiRtgOgstXYsaW2Q1By6uEVbYdARESk8/Lld7V27ryH19W2tnElN7WtrU6sIBMRERERFcIKMhEREZE+UxRoOwKdwwSZiIiISJ8JCm1HoHPYYkFEREREVAgryERERET6TMEK8ptYQSYiIiIiKoQVZCIiIiI9JrAHWQUryEREREREhTBB1pKoqCjUr19f22EQERGRvlMo1HeUU0yQi5GYmAhDQ0O0a9dO26EQERERkQYxQS7GqlWrMHz4cBw7dgxpaWnaDoeIiIhIPQSF+o5yiglyEXJycrB582Z89dVX6NChA+Lj48XHDh8+DIlEgt27d6NevXowNTWFv78/zp07J86Jj4+Hra0tEhISULNmTZiamiIwMBC3b99+63nj4uLg6ekJU1NT1K5dG4sWLVLXSyQiIiJ6RVGgvqOcYoJchE2bNqFWrVqoVasW+vTpg7i4OAiCoDRn3Lhx+P7775GUlIQqVaqgU6dOyMvLEx9/8eIFZsyYgdWrV+P48ePIzs5Gjx49ij3n8uXLMXHiRMyYMQOpqamIiYnB5MmTsXr1arW9TiIiIiJSxQS5CCtXrkSfPn0AAO3atcPz589x8OBBpTmRkZEIDAyEj48PVq9ejfv372P79u3i43l5eVi4cCECAgLQsGFDrF69GomJiTh58mSR55w2bRpmz56Nrl27wtXVFV27dsWoUaOwdOlS9b1QIiIiIrZYqOA+yG+4dOkSTp48iW3btgEAjIyM0L17d6xatQpt2rQR5wUEBIhfV6xYEbVq1UJqaqo4ZmRkBD8/P/H72rVrw9bWFqmpqWjUqJHSOR88eIDbt28jLCwMgwYNEsfz8/NhY2NTbKwymQwymUxpTCEoYCDh7z1ERERE74oJ8htWrlyJ/Px8VKtWTRwTBAHGxsbIysp663MlEslbvy9uTPH/26AsX74c/v7+So8ZGhoWe77Y2FhER0crjVWzdEZ1K5e3xklEREQkKsfbsakLS42F5OfnY82aNZg9ezZSUlLE48yZM3B2dsZPP/0kzj1x4oT4dVZWFi5fvozatWsrrXXq1Cnx+0uXLuHJkydKc16zt7dHtWrVcP36dbi7uysdrq6uxcYbERGBp0+fKh2OljX+7dtAREREpNdYQS5k165dyMrKQlhYmEprw3/+8x+sXLkSc+fOBQBMnToVdnZ2sLe3x8SJE1GpUiUEBweL842NjTF8+HDMnz8fxsbGGDZsGBo3bqzSXvFaVFQURowYAWtra7Rv3x4ymQynTp1CVlYWRo8eXeRzpFIppFKp0hjbK4iIiKg0eKtpVcymClm5ciXatGlTZN/vZ599hpSUFPz1118AgG+//RYjR45Ew4YNkZ6ejh07dsDExEScb25ujgkTJqBXr14ICAiAmZkZNm7cWOy5Bw4ciBUrViA+Ph4+Pj5o0aIF4uPj31pBJiIiIqKyJxHe3L+M3urw4cNo1aoVsrKyYGtrW+Sc+Ph4hIeH48mTJxqNDQAaO7bU+Dnf5tTDK9oOgYiISOfly+9q7dyyK4lqW1vq0URta6sTK8hERERERIWwB5mIiIhIn7EHWQVbLN4zbLEgIiIqf7TaYvH3EbWtLa3dQm1rqxNbLIiIiIiICmGLBREREZE+Y4uFClaQiYiIiIgKYQWZiIiISJ/xVtMqWEEmIiIiIiqECTIRERGRPhMU6jtKwcXFBRKJROUYOnToqzAFAVFRUXB0dISZmRlatmyJCxcuKK0hk8kwfPhwVKpUCRYWFujUqRPu3LlT6reECTIRERERaV1SUhLS09PFY//+/QCAbt26AQBmzZqFOXPmYOHChUhKSoKDgwMCAwPx7NkzcY3w8HBs374dGzduxLFjx/D8+XN06NABBQUFpYqF+yC/Z7gPMhERUfmj1X2Qz+5V29rSukHv/Nzw8HDs2rULV668yiUcHR0RHh6OCRMmAHhVLba3t8fMmTMxePBgPH36FJUrV8batWvRvXt3AMC9e/fg5OSEX375BUFBJY+FFWQiIiIiPSYIBWo7ZDIZsrOzlQ6ZTPaPMcnlcqxbtw4DBgyARCLBjRs3kJGRgbZt24pzpFIpWrRogcTERABAcnIy8vLylOY4OjrC29tbnFNS3MXiPaNrFVtHy4raDkF07/ljbYdARESkV2JjYxEdHa00FhkZiaioqLc+LyEhAU+ePEFoaCgAICMjAwBgb2+vNM/e3h63bt0S55iYmKBChQoqc14/v6SYIBMRERHpMzXeKCQiIgKjR49WGpNKpf/4vJUrV6J9+/ZwdHRUGpdIJErfC4KgMvamksx5E1ssiIiIiEgtpFIprK2tlY5/SpBv3bqFAwcOYODAgeKYg4MDAKhUgjMzM8WqsoODA+RyObKysoqdU1JMkImIiIj0mUKhvuMdxMXFoUqVKvj000/FMVdXVzg4OIg7WwCv+pSPHDmCJk2aAAAaNmwIY2NjpTnp6ek4f/68OKek2GJBRERERDpBoVAgLi4O/fr1g5HR/9JUiUSC8PBwxMTEwMPDAx4eHoiJiYG5uTl69eoFALCxsUFYWBjGjBkDOzs7VKxYEWPHjoWPjw/atGlTqjiYIBMRERHpMzX2IJfWgQMHkJaWhgEDBqg8Nn78eOTm5mLIkCHIysqCv78/9u3bBysrK3HO3LlzYWRkhJCQEOTm5qJ169aIj4+HoaFhqeLgPsjvGSOTatoOQQl3sSAiIvpn2twH+WVygtrWNm0YrLa11YkVZCIiIiJ9pijdXeb0ARNkIiIiIn2mQy0WuqLc72Jx8+ZNSCQSpKSklOm6EokECQkJZbomEREREek+nU+QQ0NDIZFIxMPOzg7t2rXD2bNntR0aERERUfmnY9u86QKdT5ABoF27dkhPT0d6ejoOHjwIIyMjdOjQQdthlZpcLi9yPC8vT8OREBEREVFxykWCLJVK4eDgAAcHB9SvXx8TJkzA7du38eDBA5W58fHxsLW1VRpLSEhQucXgzp070bBhQ5iamsLNzQ3R0dHIz89XmpOeno727dvDzMwMrq6u+Pnnn5Uev3v3Lrp3744KFSrAzs4OnTt3xs2bN8XHQ0NDERwcjNjYWDg6OqJmzZpiS8jmzZvRsmVLmJqaYtmyZbC2tsaWLVtUYrSwsMCzZ8/e4V0jIiIiKgFBob6jnCoXCXJhz58/x08//QR3d3fY2dm90xp79+5Fnz59MGLECFy8eBFLly5FfHw8ZsyYoTRv8uTJ+Oyzz3DmzBn06dMHPXv2RGpqKgDgxYsXaNWqFSwtLfH777/j2LFjsLS0RLt27ZQqxQcPHkRqair279+PXbt2ieMTJkzAiBEjkJqaii5duqBHjx6Ii4tTOn9cXBz+85//KO3vR0RERETqVS52sdi1axcsLS0BADk5OahatSp27doFA4N3y+9nzJiBr7/+Gv369QMAuLm5Ydq0aRg/fjwiIyPFed26dRPvAz5t2jTs378fCxYswKJFi7Bx40YYGBhgxYoVYnU6Li4Otra2OHz4MNq2bQsAsLCwwIoVK2BiYgIAYoU5PDwcXbt2Fc81cOBANGnSBPfu3YOjoyMePnyIXbt2Kd0ukYiIiKjMleNeYXUpFwlyq1atsHjxYgDA48ePsWjRIrRv3x4nT558p/WSk5ORlJSkVDEuKCjAy5cv8eLFC5ibmwMAAgIClJ4XEBAg7paRnJyMq1evqlR3X758iWvXronf+/j4iMlxYX5+fkrfN2rUCHXq1MGaNWvw9ddfY+3atahRowaaN29e7OuQyWSQyWRKY4IgqLSTEBEREVHJlYsE2cLCAu7u7uL3DRs2hI2NDZYvXy5WeF8zMDDAmzcHfPMiOIVCgejoaKUK7mumpqZvjeV18qlQKNCwYUP89NNPKnMqV66sFHtxr+lNAwcOxMKFC/H1118jLi4O/fv3f2uyGxsbi+joaOX4DCwhMbR+62sgIiIiErGCrKJcJMhvkkgkMDAwQG5urspjlStXxrNnz5CTkyMmoW/ukdygQQNcunRJKekuyokTJ9C3b1+l7319fcU1Nm3ahCpVqsDaumwS0j59+mD8+PGYP38+Lly4ILaAFCciIgKjR49WGqtgV7tMYiEiIiL9IAi8k96bykWCLJPJkJGRAQDIysrCwoUL8fz5c3Ts2FFlrr+/P8zNzfHNN99g+PDhOHnyJOLj45XmTJkyBR06dICTkxO6desGAwMDnD17FufOncP06dPFeT///DP8/PzQrFkz/PTTTzh58iRWrlwJAOjduze+++47dO7cGVOnTkX16tWRlpaGbdu2Ydy4cahevXqpX2eFChXQtWtXjBs3Dm3btv3HNaRSKaRSqdIY2yuIiIiI/p1ysYvFnj17ULVqVVStWhX+/v5ISkrCzz//jJYtW6rMrVixItatW4dffvkFPj4+2LBhA6KiopTmBAUFiRfAffjhh2jcuDHmzJkDZ2dnpXnR0dHYuHEj6tati9WrV+Onn36Cl5cXAMDc3By///47atSoga5du8LT0xMDBgxAbm7uv6ooh4WFQS6XY8CAAe+8BhEREVGJ8UYhKiTCmw27pFU//fQTRo4ciXv37hV5cd8/MTKppoao3p2jZUVthyC69/yxtkMgIiIqUr78rtbOnXt4ldrWNmtZPgt+5aLFQh+8ePECN27cQGxsLAYPHvxOyTERERFRqZXjG3qoS7losdAHs2bNQv369WFvb4+IiAhth0NERESkt9hi8Z5hi0Xx2GJBRES6SqstFgeXqW1ts9ZfqG1tdWIFmYiIiIioEPYgExEREekz9iCrYIJMREREpM/K8XZs6sIWCyIiIiKiQlhBJiIiItJnbLFQwQoyEREREVEhrCATERER6TP2IKtgBZmIiIiIqBBWkImIiIj0GSvIKpggk1rp0t3rqlnZaTsE0d1nj7QdAhERERWDCTIRERGRPuMuFiqYIBMRERHpM7ZYqOBFekREREREhbCCTERERKTP2GKhghVkIiIiIqJCWEEmIiIi0mfsQVbBCjIRERERUSGsIBMRERHpM/Ygq2AF+S0kEgkSEhK0HQYRERERaZBeJ8gZGRkYPnw43NzcIJVK4eTkhI4dO+LgwYPaDo2IiIhIMxQK9R3llN62WNy8eRNNmzaFra0tZs2ahbp16yIvLw979+7F0KFD8ffff6vlvHK5HCYmJkpjgiCgoKAARkZ6+8dBRERE2lKOE1l10dsK8pAhQyCRSHDy5En85z//Qc2aNVGnTh2MHj0aJ06cEOc9fPgQXbp0gbm5OTw8PLBjxw6ldS5evIhPPvkElpaWsLe3x+eff46HDx+Kj7ds2RLDhg3D6NGjUalSJQQGBuLw4cOQSCTYu3cv/Pz8IJVKsXbtWhgYGODUqVNK6y9YsADOzs4QBEG9bwgRERERAdDTBPnx48fYs2cPhg4dCgsLC5XHbW1txa+jo6MREhKCs2fP4pNPPkHv3r3x+PFjAEB6ejpatGiB+vXr49SpU9izZw/u37+PkJAQpfVWr14NIyMjHD9+HEuXLhXHx48fj9jYWKSmpqJTp05o06YN4uLilJ4bFxeH0NBQSCSSMnwHiIiIiP6fIKjvKKf0MkG+evUqBEFA7dq1/3FuaGgoevbsCXd3d8TExCAnJwcnT54EACxevBgNGjRATEwMateuDV9fX6xatQq//fYbLl++LK7h7u6OWbNmoVatWkrnnDp1KgIDA/HBBx/Azs4OAwcOxIYNGyCTyQAAZ86cQUpKCvr371/G7wARERERFUcvE+TX7QolqcrWrVtX/NrCwgJWVlbIzMwEACQnJ+O3336DpaWleLxOgK9duyY+z8/Pr8i13xwPDg6GkZERtm/fDgBYtWoVWrVqBRcXlyKfL5PJkJ2drXSwFYOIiIhKhRfpqdDLBNnDwwMSiQSpqan/ONfY2Fjpe4lEAsX//4ErFAp07NgRKSkpSseVK1fQvHlz8TlFtXEUNW5iYoLPP/8ccXFxkMvlWL9+PQYMGFBsbLGxsbCxsVE6BMWzf3xNRERERFQ8vUyQK1asiKCgIPz444/IyclRefzJkyclWqdBgwa4cOECXFxc4O7urnQUlxT/k4EDB+LAgQNYtGgR8vLy0LVr12LnRkRE4OnTp0qHxMDqnc5LREREeooVZBV6mSADwKJFi1BQUIBGjRph69atuHLlClJTUzF//nwEBASUaI2hQ4fi8ePH6NmzJ06ePInr169j3759GDBgAAoKCt4pLk9PTzRu3BgTJkxAz549YWZmVuxcqVQKa2trpYMX8xERERH9O3qbILu6uuKvv/5Cq1atMGbMGHh7eyMwMBAHDx7E4sWLS7SGo6Mjjh8/joKCAgQFBcHb2xsjR46EjY0NDAze/a0NCwuDXC5/a3sFERERUZkQFOo7yimJwKu6dM6MGTOwceNGnDt3rtTPNTKppoaI3g/VrOy0HYLo7rNH2g6BiIh0SL78rtbOnbsmQm1rm/WNVdva6qS3FWRd9Pz5cyQlJWHBggUYMWKEtsMhIiIi0ktMkHXIsGHD0KxZM7Ro0YLtFURERKQZvFGICiNtB0D/Ex8fj/j4eG2HQURERKTXWEEmIiIi0mc6tM3b3bt30adPH9jZ2cHc3Bz169dHcnKy+LggCIiKioKjoyPMzMzQsmVLXLhwQWkNmUyG4cOHo1KlSrCwsECnTp1w586dUsXBBJmIiIiItC4rKwtNmzaFsbExfv31V1y8eBGzZ8+Gra2tOGfWrFmYM2cOFi5ciKSkJDg4OCAwMBDPnv3vRmnh4eHYvn07Nm7ciGPHjuH58+fo0KFDqbbg5S4W7xnuYlE87mJBRES6Squ7WKwcq7a1zcK+L/Hcr7/+GsePH8fRo0eLfFwQBDg6OiI8PBwTJkwA8KpabG9vj5kzZ2Lw4MF4+vQpKleujLVr16J79+4AgHv37sHJyQm//PILgoKCShQLK8hEREREpBYymQzZ2dlKh0wmK3Lujh074Ofnh27duqFKlSrw9fXF8uXLxcdv3LiBjIwMtG3bVhyTSqVo0aIFEhMTAQDJycnIy8tTmuPo6Ahvb29xTkkwQSYiIiLSZ2q8UUhsbCxsbGyUjtjYovdGvn79OhYvXgwPDw/s3bsXX375JUaMGIE1a9YAADIyMgAA9vb2Ss+zt7cXH8vIyICJiQkqVKhQ7JyS4C4WRERERHpMUKiv2zYiIgKjR49WGpNKpUXOVSgU8PPzQ0xMDADA19cXFy5cwOLFi9G3b19xnkQiUXqeIAgqY28qyZzCWEEmIiIiIrWQSqWwtrZWOopLkKtWrQovLy+lMU9PT6SlpQEAHBwcAEClEpyZmSlWlR0cHCCXy5GVlVXsnJJggkxERESkz3Rkm7emTZvi0qVLSmOXL1+Gs7MzAMDV1RUODg7Yv3+/+LhcLseRI0fQpEkTAEDDhg1hbGysNCc9PR3nz58X55QEWyyIiIiISOtGjRqFJk2aICYmBiEhITh58iSWLVuGZcuWAXjVWhEeHo6YmBh4eHjAw8MDMTExMDc3R69evQAANjY2CAsLw5gxY2BnZ4eKFSti7Nix8PHxQZs2bUocCxNk0hu6tLVat6ofajsEJT+nJ2k7BFHJO8Q0g/tgUnmnSz9TuvbzpEvvjVYJpb+hhzp8+OGH2L59OyIiIjB16lS4urpi3rx56N27tzhn/PjxyM3NxZAhQ5CVlQV/f3/s27cPVlZW4py5c+fCyMgIISEhyM3NRevWrREfHw9DQ8MSx8J9kN8z3Ae5fGCCXDxd+weL/4Ok8k6XfqZ07edJl96bPC3ug/xi8XC1rW3+1QK1ra1OrCATERER6TM17mJRXvEiPSIiIiKiQlhBJiIiItJnpdxtQh+wgkxEREREVAgryERERET6jBVkFUyQiYiIiPQZNzRTwRYLIiIiIqJCWEEmIiIi0mdssVDBCvI7yMjIwPDhw+Hm5gapVAonJyd07NgRBw8e1HZoRERERPQvsYJcSjdv3kTTpk1ha2uLWbNmoW7dusjLy8PevXsxdOhQ/P3336VeMy8vD8bGxmqIloiIiOgf8EYhKlhBLqUhQ4ZAIpHg5MmT+M9//oOaNWuiTp06GD16NE6cOAEASEtLQ+fOnWFpaQlra2uEhITg/v374hpRUVGoX78+Vq1aJVahBUGARCLBihUr0KVLF5ibm8PDwwM7duzQ1kslIiIi0ktMkEvh8ePH2LNnD4YOHQoLCwuVx21tbSEIAoKDg/H48WMcOXIE+/fvx7Vr19C9e3eluVevXsXmzZuxdetWpKSkiOPR0dEICQnB2bNn8cknn6B37954/Pixul8aERER6StBob6jnGKLRSlcvXoVgiCgdu3axc45cOAAzp49ixs3bsDJyQkAsHbtWtSpUwdJSUn48MMPAQByuRxr165F5cqVlZ4fGhqKnj17AgBiYmKwYMECnDx5Eu3atVPTqyIiIiKiwpggl4Lw//sESiSSYuekpqbCyclJTI4BwMvLC7a2tkhNTRUTZGdnZ5XkGADq1q0rfm1hYQErKytkZmYWeS6ZTAaZTKYS49viIyIiIlLCHmQVbLEoBQ8PD0gkEqSmphY7p7gE9c3xolo0AKhcrCeRSKAoZvuV2NhY2NjYKB2C4llJXgoRERERAEBQKNR2lFdMkEuhYsWKCAoKwo8//oicnByVx588eQIvLy+kpaXh9u3b4vjFixfx9OlTeHp6lmk8ERERePr0qdIhMbAq03MQERER6RsmyKW0aNEiFBQUoFGjRti6dSuuXLmC1NRUzJ8/HwEBAWjTpg3q1q2L3r1746+//sLJkyfRt29ftGjRAn5+fmUai1QqhbW1tdLB9goiIiIqFYWgvqOcYoJcSq6urvjrr7/QqlUrjBkzBt7e3ggMDMTBgwexePFiSCQSJCQkoEKFCmjevDnatGkDNzc3bNq0SduhExEREVEJSITXV57Re8HIpJq2Q6AS6Fb1Q22HoOTn9CRthyDStc9A+D9IKu906WdK136edOm9yZPf1dq5c6b3UdvaFpPWqW1tdWIFmYiIiIioEG7zRkRERKTPynGvsLqwgkxEREREVAgryERERET6rBzvV6wuTJCJiIiI9BlbLFSwxYKIiIiIqBBWkImIiIj0mcAWizexgkxEREREVAgryERERET6jD3IKlhBJiIiIiIqhBVkIiIiIj0mcJs3FawgExEREREVwgrye8bQQLd+5ynQod9KDSQSbYcg+jk9SdshKBnk2FTbIYiW3zuu7RCI3iu61F2qa/9GSQ2NtR2CbmAPsgomyERERET6jAmyCt36VY6IiIiISMtYQSYiIiLSZ7xRiApWkImIiIiICmEFmYiIiEifsQdZBSvIRERERESFsIJMREREpMcEVpBVsIJMRERERFQIK8hERERE+owVZBWsIBeSkZGBkSNHwt3dHaamprC3t0ezZs2wZMkSvHjxQtvhEREREZU9hUJ9RznFCvL/u379Opo2bQpbW1vExMTAx8cH+fn5uHz5MlatWgVHR0d06tRJ5Xl5eXkwNuatKomIiIjeF6wg/78hQ4bAyMgIp06dQkhICDw9PeHj44PPPvsMu3fvRseOHQEAEokES5YsQefOnWFhYYHp06ejoKAAYWFhcHV1hZmZGWrVqoUffvhBaf3Q0FAEBwcjOjoaVapUgbW1NQYPHgy5XC7OEQQBs2bNgpubG8zMzFCvXj1s2bJFo+8DERER6RmFoL6jnGIFGcCjR4+wb98+xMTEwMLCosg5EolE/DoyMhKxsbGYO3cuDA0NoVAoUL16dWzevBmVKlVCYmIivvjiC1StWhUhISHi8w4ePAhTU1P89ttvuHnzJvr3749KlSphxowZAIBJkyZh27ZtWLx4MTw8PPD777+jT58+qFy5Mlq0aKHeN4GIiIiIADBBBgBcvXoVgiCgVq1aSuOVKlXCy5cvAQBDhw7FzJkzAQC9evXCgAEDlOZGR0eLX7u6uiIxMRGbN29WSpBNTEywatUqmJubo06dOpg6dSrGjRuHadOmITc3F3PmzMGhQ4cQEBAAAHBzc8OxY8ewdOlSJshERESkHuW40qsuTJALKVwlBoCTJ09CoVCgd+/ekMlk4rifn5/Kc5csWYIVK1bg1q1byM3NhVwuR/369ZXm1KtXD+bm5uL3AQEBeP78OW7fvo3MzEy8fPkSgYGBSs+Ry+Xw9fUtMl6ZTKYUF/CqTePN10FEREREJccEGYC7uzskEgn+/vtvpXE3NzcAgJmZmdL4m20YmzdvxqhRozB79mwEBATAysoK3333Hf78888SnV8ikUDx/1d67t69G9WqVVN6XCqVFvm82NhYpco1ABgYWsHIyKZE5yUiIiISBFaQ38QEGYCdnR0CAwOxcOFCDB8+vNg+5OIcPXoUTZo0wZAhQ8Sxa9euqcw7c+YMcnNzxYT7xIkTsLS0RPXq1VGhQgVIpVKkpaWVuJ0iIiICo0ePVhqrVNmrVLETERERkTImyP9v0aJFaNq0Kfz8/BAVFYW6devCwMAASUlJ+Pvvv9GwYcNin+vu7o41a9Zg7969cHV1xdq1a5GUlARXV1eleXK5HGFhYZg0aRJu3bqFyMhIDBs2DAYGBrCyssLYsWMxatQoKBQKNGvWDNnZ2UhMTISlpSX69euncl6pVKpSXWZ7BREREZUKe5BVcJu3//fBBx/g9OnTaNOmDSIiIlCvXj34+flhwYIFGDt2LKZNm1bsc7/88kt07doV3bt3h7+/Px49eqRUTX6tdevW8PDwQPPmzRESEoKOHTsiKipKfHzatGmYMmUKYmNj4enpiaCgIOzcuVMl0SYiIiIqMzqyzVtUVBQkEonS4eDgID4uCAKioqLg6OgIMzMztGzZEhcuXFBaQyaTYfjw4ahUqRIsLCzQqVMn3Llzp9RviURg44lGhIaG4smTJ0hISFDreaSmTmpdv7QKdOguOgY6VF1X6NiP3SDHptoOQbT83nFth0BEamJooFt1Oamh7tzoKzvnuvbOHRb4z5PekfXK/SWeGxUVhS1btuDAgQPimKGhISpXrgwAmDlzJmbMmIH4+HjUrFkT06dPx++//45Lly7BysoKAPDVV19h586diI+Ph52dHcaMGYPHjx8jOTkZhoaGJY6FLRZEREREekzQoRYLIyMjparxa4IgYN68eZg4cSK6du0KAFi9ejXs7e2xfv16DB48GE+fPsXKlSuxdu1atGnTBgCwbt06ODk54cCBAwgKCipxHLr1qxwRERERvTdkMhmys7OVjje3qC3sypUrcHR0hKurK3r06IHr119V1m/cuIGMjAy0bdtWnCuVStGiRQskJiYCAJKTk5GXl6c0x9HREd7e3uKckmKCrCHx8fFqb68gIiIiKjU19iDHxsbCxsZG6YiNjS0yDH9/f3HTg+XLlyMjIwNNmjTBo0ePkJGRAQCwt7dXeo69vb34WEZGBkxMTFChQoVi55QUWyyIiIiISC2K2pK2uPs7tG/fXvzax8cHAQEB+OCDD7B69Wo0btwYgOpuXSW5Qdq73ESNFWQiIiIifaZQ3yGVSmFtba10FJcgv8nCwgI+Pj64cuWK2Jf8ZiU4MzNTrCo7ODhALpcjKyur2DklxQSZiIiIiHSOTCZDamoqqlatCldXVzg4OGD//v/tiiGXy3HkyBE0adIEANCwYUMYGxsrzUlPT8f58+fFOSXFFgsiIiIiPaYru1iMHTsWHTt2RI0aNZCZmYnp06cjOzsb/fr1g0QiQXh4OGJiYuDh4QEPDw/ExMTA3NwcvXr1AgDY2NggLCwMY8aMgZ2dHSpWrIixY8fCx8dH3NWipJggExEREekzHUmQ79y5g549e+Lhw4eoXLkyGjdujBMnTsDZ2RkAMH78eOTm5mLIkCHIysqCv78/9u3bJ+6BDABz586FkZERQkJCkJubi9atWyM+Pr5UeyADvFHIe4c3CikebxRSPN4ohIg0gTcKKZ42bxTypGcrta1tu+E3ta2tTqwgExEREekz3all6Qzd+lWOiIiIiEjLWEEmIiIi0mO6cpGeLmGC/J7RpZ5fXaNrfb+6RJf6fsc4Ntd2CEpm3/td2yEQvTd07d+oXEXxtzwm/cYEmYiIiEif6dbvLTqBPchERERERIWwgkxERESkx9iDrIoJMhEREZE+Y4uFCrZYEBEREREVwgoyERERkR4TWEFWwQoyEREREVEhrCATERER6TNWkFWwgkxEREREVIjeJMi3b99GWFgYHB0dYWJiAmdnZ4wcORKPHj3SdmhEREREWiMo1HeUV2WSID958qQsllGb69evw8/PD5cvX8aGDRtw9epVLFmyBAcPHkRAQAAeP35c5PPkcrmGIyUiIiIibSt1gjxz5kxs2rRJ/D4kJAR2dnaoVq0azpw5U6bBlZWhQ4fCxMQE+/btQ4sWLVCjRg20b98eBw4cwN27dzFx4kQAgIuLC6ZPn47Q0FDY2Nhg0KBBAIAJEyagZs2aMDc3h5ubGyZPnoy8vDxx/aioKNSvXx9r166Fi4sLbGxs0KNHDzx79kyc8+zZM/Tu3RsWFhaoWrUq5s6di5YtWyI8PFycI5fLMX78eFSrVg0WFhbw9/fH4cOHNfIeERERkZ5SqPEop0qdIC9duhROTk4AgP3792P//v349ddf0b59e4wbN67MA/y3Hj9+jL1792LIkCEwMzNTeszBwQG9e/fGpk2bIAiv7iLz3XffwdvbG8nJyZg8eTIAwMrKCvHx8bh48SJ++OEHLF++HHPnzlVa69q1a0hISMCuXbuwa9cuHDlyBN9++634+OjRo3H8+HHs2LED+/fvx9GjR/HXX38prdG/f38cP34cGzduxNmzZ9GtWze0a9cOV65cUcdbQ0RERERFKPUuFunp6WKCvGvXLoSEhKBt27ZwcXGBv79/mQf4b125cgWCIMDT07PIxz09PZGVlYUHDx4AAD7++GOMHTtWac6kSZPEr11cXDBmzBhs2rQJ48ePF8cVCgXi4+NhZWUFAPj8889x8OBBzJgxA8+ePcPq1auxfv16tG7dGgAQFxcHR0dH8fnXrl3Dhg0bcOfOHXF87Nix2LNnD+Li4hATE1MG7wYRERGRsvLcK6wupU6QK1SogNu3b8PJyQl79uzB9OnTAQCCIKCgoKDMA1S315VjiUQCAPDz81OZs2XLFsybNw9Xr17F8+fPkZ+fD2tra6U5Li4uYnIMAFWrVkVmZiaAVz3QeXl5aNSokfi4jY0NatWqJX7/119/QRAE1KxZU2ldmUwGOzu7ImOXyWSQyWQqr+f1ayEiIiL6J0yQVZU6Qe7atSt69eoFDw8PPHr0CO3btwcApKSkwN3dvcwD/Lfc3d0hkUhw8eJFBAcHqzz+999/o0KFCqhUqRIAwMLCQunxEydOoEePHoiOjkZQUBBsbGywceNGzJ49W2mesbGx0vcSiQQKxau/cW8m4a+9HgdeVaANDQ2RnJwMQ0NDpXmWlpZFvrbY2FhER0crn9fAEhJD6yLnExEREdE/K3UP8ty5czFs2DB4eXlh//79YvKWnp6OIUOGlHmA/5adnR0CAwOxaNEi5ObmKj2WkZGBn376Cd27dy+26nr8+HE4Oztj4sSJ8PPzg4eHB27dulWqGD744AMYGxvj5MmT4lh2drZSb7Gvry8KCgqQmZkJd3d3pcPBwaHIdSMiIvD06VOlQ2JgVeRcIiIioqJwmzdVpa4gGxsbq/ToAlDajUHXLFy4EE2aNEFQUBCmT58OV1dXXLhwAePGjUO1atUwY8aMYp/r7u6OtLQ0bNy4ER9++CF2796N7du3l+r8VlZW6NevH8aNG4eKFSuiSpUqiIyMhIGBgZiY16xZE71790bfvn0xe/Zs+Pr64uHDhzh06BB8fHzwySefqKwrlUohlUqVxtheQURERPTvlChB3rFjR4kX7NSp0zsHoy4eHh44deoUoqKi0L17dzx69AgODg4IDg5GZGQkKlasWOxzO3fujFGjRmHYsGGQyWT49NNPMXnyZERFRZUqhjlz5uDLL79Ehw4dYG1tjfHjx+P27dswNTUV58TFxWH69OkYM2YM7t69Czs7OwQEBBSZHBMRERGVCYHFtTdJhMKNsMUwMChZJ4ZEIimXF+ppQ05ODqpVq4bZs2cjLCyszNY1MqlWZmsRacMYx+baDkHJ7Hu/azsEIlITXUoL8+R3tXbu+y1bqm1t+3J6P4cSVZBfX2xG7+706dP4+++/0ahRIzx9+hRTp04F8KpCTURERKQt5blXWF1K3YNc2MuXL5VaBOjtvv/+e1y6dAkmJiZo2LAhjh49Ku6eQURERES6odQJckFBAWJiYrBkyRLcv38fly9fFm+/7OLiUqbtAu8TX19fJCcnazsMIiIiIiWCQpeaTXRDqbd5mzFjBuLj4zFr1iyYmJiI4z4+PlixYkWZBkdERERE6sVt3lSVOkFes2YNli1bht69eyvd0KJu3br4+++/yzQ4IiIiIiJNK3WLxd27d4u8Y55CoUBeXl6ZBEVEREREmiFwmzcVpa4g16lTB0ePHlUZ//nnn+Hr61smQRERERERaUupK8iRkZH4/PPPcffuXSgUCmzbtg2XLl3CmjVrsGvXLnXESERERERqUp57hdWl1BXkjh07YtOmTfjll18gkUgwZcoUpKamYufOnQgMDFRHjEREREREGvNO+yAHBQUhKCiorGMhIiIiIg3jNm+q3vlGIadOnUJqaiokEgk8PT3RsGHDsoyLiIiIiEgrSp0g37lzBz179sTx48dha2sLAHjy5AmaNGmCDRs2wMnJqaxjJCIiIiI1EQRtR6B7Sp0gDxgwAHl5eUhNTUWtWrUAAJcuXcKAAQMQFhaGffv2lXmQVHK69iEJf+aotObc+13bIShpae+t7RBEh++f13YIRO8V/hv1ClssVJU6QT569CgSExPF5BgAatWqhQULFqBp06ZlGhwRERERkaaVOkGuUaNGkTcEyc/PR7Vq1cokKCIiIiLSDFaQVZV6m7dZs2Zh+PDhOHXqFIT/b1o5deoURo4cie+//77MAyQiIiIi0qQSVZArVKgAieR/v13k5OTA398fRkavnp6fnw8jIyMMGDAAwcHBagmUiIiIiMoeL9JTVaIEed68eWoOg4iIiIhIN5QoQe7Xr5+64yAiIiIiLWAPsqp3vlEIAOTm5qpcsGdtbf2vAiIiIiIi0qZSJ8g5OTmYMGECNm/ejEePHqk8XlBQUCaBEREREZH6CQIryG8q9S4W48ePx6FDh7Bo0SJIpVKsWLEC0dHRcHR0xJo1a9QRo16QSCRISEjQdhhERESkZwSF+o7yqtQJ8s6dO7Fo0SL85z//gZGRET766CNMmjQJMTEx+Omnn9QRo0aEhoZCIpFAIpHA2NgYbm5uGDt2LHJycrQdGhERERFpUKlbLB4/fgxXV1cAr/qNHz9+DABo1qwZvvrqq7KNTsPatWuHuLg45OXl4ejRoxg4cCBycnKwePFibYdGREREpBYKtlioKHUF2c3NDTdv3gQAeHl5YfPmzQBeVZZtbW3LMjaNk0qlcHBwgJOTE3r16oXevXsjISEBgiBg1qxZcHNzg5mZGerVq4ctW7aIzysoKEBYWBhcXV1hZmaGWrVq4YcfflBZf9WqVahTpw6kUimqVq2KYcOGKT3+8OFDdOnSBebm5vDw8MCOHTvU/pqJiIiISFmpE+T+/fvjzJkzAICIiAixF3nUqFEYN25cmQeoTWZmZsjLy8OkSZMQFxeHxYsX48KFCxg1ahT69OmDI0eOAAAUCgWqV6+OzZs34+LFi5gyZQq++eYb8ZcHAFi8eDGGDh2KL774AufOncOOHTvg7u6udL7o6GiEhITg7Nmz+OSTT9C7d2+xQk9ERESkDoIgUdtRXkkE4d/dPyUtLQ2nTp3CBx98gHr16pVVXBoXGhqKJ0+eiBfKnTx5Ep988glatWqFXbt24dChQwgICBDnDxw4EC9evMD69euLXG/o0KG4f/++WGmuVq0a+vfvj+nTpxc5XyKRYNKkSZg2bRqAV7uFWFlZ4ZdffkG7du1K/DqMTaqVeK4m8OY8VFq69r/TFvbe2g5BdPj+eW2HQERqki+/q7VzX6rdXm1r1/r713d+bmxsLL755huMHDlSvGmdIAiIjo7GsmXLkJWVBX9/f/z444+oU6eO+DyZTIaxY8diw4YNyM3NRevWrbFo0SJUr169xOf+V/sgA0CNGjVQo0aNf7uMTti1axcsLS2Rn5+PvLw8dO7cGWPHjsWWLVsQGBioNFcul8PX11f8fsmSJVixYgVu3bqF3NxcyOVy1K9fHwCQmZmJe/fuoXXr1m89f926dcWvLSwsYGVlhczMzGLny2QyyGQypTFBEJRuC05ERET0Nrp4o5CkpCQsW7ZMKTcCgFmzZmHOnDmIj49HzZo1MX36dAQGBuLSpUuwsrICAISHh2Pnzp3YuHEj7OzsMGbMGHTo0AHJyckwNDQs0flLlCDPnz+/xC9oxIgRJZ6ra1q1aoXFixfD2NgYjo6OMDY2xp9//gkA2L17N6pVU67OSqVSAMDmzZsxatQozJ49GwEBAbCyssJ3330nPtfMzKxE5zc2Nlb6XiKRQKEofo+U2NhYREdHKz/HwBKGhrxZCxEREZVPz58/R+/evbF8+XKlT94FQcC8efMwceJEdO3aFQCwevVq2NvbY/369Rg8eDCePn2KlStXYu3atWjTpg0AYN26dXBycsKBAwcQFBRUohhKlCDPnTu3RItJJJJynSBbWFio9AV7eXlBKpUiLS0NLVq0KPJ5R48eRZMmTTBkyBBx7Nq1a+LXVlZWcHFxwcGDB9GqVasyizciIgKjR49WGqtoV7vM1iciIqL3379rti17Q4cOxaeffoo2bdooJcg3btxARkYG2rZtK45JpVK0aNECiYmJGDx4MJKTk5GXl6c0x9HREd7e3khMTCzbBPnGjRslfU3vHSsrK4wdOxajRo2CQqFAs2bNkJ2djcTERFhaWqJfv35wd3fHmjVrsHfvXri6umLt2rVISkoSt8MDgKioKHz55ZeoUqUK2rdvj2fPnuH48eMYPnz4O8cmlUrFKvZrbK8gIiKi0lBni0VR7aBF5S+vbdy4EX/99ReSkpJUHsvIyAAA2NvbK43b29vj1q1b4hwTExNUqFBBZc7r55dEqXex0EfTpk3DlClTEBsbC09PTwQFBWHnzp1iAvzll1+ia9eu6N69O/z9/fHo0SOlajIA9OvXD/PmzcOiRYtQp04ddOjQAVeuXNHGyyEiIiLSiNjYWNjY2CgdsbGxRc69ffs2Ro4ciXXr1sHU1LTYNd8sBpbk+qvSXqP1r3exIN3CXSyovNO1z0C4iwURaYI2d7E479ZBbWt7pG4tcQU5ISEBXbp0UbqQrqCgABKJBAYGBrh06RLc3d3x119/KW2U0LlzZ9ja2mL16tU4dOgQWrdujcePHytVkevVq4fg4GCVa7eKwwoyEREREamFVCqFtbW10lFce0Xr1q1x7tw5pKSkiIefnx969+6NlJQUuLm5wcHBAfv37xefI5fLceTIETRp0gQA0LBhQxgbGyvNSU9Px/nz58U5JfGvt3kjIiIiovJLV27oYWVlBW9v5U/tLCwsYGdnJ46Hh4cjJiYGHh4e8PDwQExMDMzNzdGrVy8AgI2NDcLCwjBmzBjY2dmhYsWKGDt2LHx8fMRdLUqCCTIRERERlQvjx49Hbm4uhgwZIt4oZN++feIeyMCr3deMjIwQEhIi3igkPj6+xHsgAyXsQT579myJF3xzQ2fSLPYgU3mnG3WM/2EPMhFpgjZ7kM+6dFTb2nVv7lTb2upUogpy/fr1IZFISnQFYEFBQZkERkRERESkDaXeB/n06dMYO3Ysxo0bh4CAAADAH3/8gdmzZ2PWrFnqiZKIiIiI1EKhIz3IuqRECbKzs7P4dbdu3TB//nx88skn4ljdunXh5OSEyZMnIzg4uMyDJCIiIiL10JWL9HRJqbd5O3funNId4l5zdXXFxYsXyyQoIiIiIiJtKXWC7OnpienTp+Ply5fimEwmw/Tp0+Hp6VmmwRERERGRegmC+o7yqtTbvC1ZsgQdO3aEk5MT6tWrBwA4c+YMJBIJdu3aVeYBEhERERFpUqkT5EaNGuHGjRtYt24d/v77bwiCgO7du6NXr16wsLBQR4xEREREpCa8SE/VO90oxNzcHF988UVZx0JEREREpHXvlCCvXbsWS5cuxfXr1/HHH3/A2dkZc+fOhZubGzp37lzWMVIpGBiUuq1crUpwHxqN+ac9vDWpQKHQdgg6S5f+nADgiA7dnMOvkoe2Q1By6uEVbYdA5Yxu/XTTa9zFQlWps6nFixdj9OjRaN++PbKyssQbg1SoUAHz5s0r6/iIiIiIiDSq1AnyggULsHz5ckycOBFGRv8rQPv5+eHcuXNlGhwRERERqZdCkKjtKK9K3WJx48YN+Pr6qoxLpVLk5OSUSVBEREREpBm60wypO0pdQXZ1dUVKSorK+K+//govL6+yiImIiIiISGtKXUEeN24chg4dipcvX0IQBJw8eRIbNmxAbGwsVqxYoY4YiYiIiEhNynMrhLqUOkHu378/8vPzMX78eLx48QK9evVCtWrV8MMPP6BHjx7qiJGIiIiISGPeaZu3QYMGYdCgQXj48CEUCgWqVKlS1nERERERkQZwmzdVpe5B/vjjj/HkyRMAQKVKlcTkODs7Gx9//HGZBkdEREREpGmlriAfPnwYcrlcZfzly5c4evRomQRFRERERJrBW1epKnGCfPbsWfHrixcvIiMjQ/y+oKAAe/bsQbVq1co2ujIWGhqK1atXAwCMjIzg5OSErl27Ijo6GhYWFlqOjoiIiIh0QYkT5Pr160MikUAikRTZSmFmZoYFCxaUaXDq0K5dO8TFxSEvLw9Hjx7FwIEDkZOTg8WLF2s7tGLl5eXB2NhY22EQERHRe0jgTcBVlLgH+caNG7h27Zq4tduNGzfE4+7du8jOzsaAAQPUGWuZkEqlcHBwgJOTE3r16oXevXsjISEBgiBg1qxZcHNzg5mZGerVq4ctW7aIzzt8+DAkEgl2796NevXqwdTUFP7+/kp3D4yPj4etrS0SEhJQs2ZNmJqaIjAwELdv31aKYefOnWjYsCFMTU3h5uaG6Oho5Ofni49LJBIsWbIEnTt3hoWFBaZPn67+N4aIiIj0kkJQ31FelbiC7OzsDABQKN6vThUzMzPk5eVh0qRJ2LZtGxYvXgwPDw/8/vvv6NOnDypXrowWLVqI88eNG4cffvgBDg4O+Oabb9CpUydcvnxZrPC+ePECM2bMwOrVq2FiYoIhQ4agR48eOH78OABg79696NOnD+bPn4+PPvoI165dwxdffAEAiIyMFM8TGRmJ2NhYzJ07F4aGhhp8R4iIiIj0W6kv0ouNjYW9vb1KtXjVqlV48OABJkyYUGbBqdvJkyexfv16tGrVCnPmzMGhQ4cQEBAAAHBzc8OxY8ewdOlSpQQ5MjISgYGBAIDVq1ejevXq2L59O0JCQgC8aodYuHAh/P39xTmenp44efIkGjVqhBkzZuDrr79Gv379xPNMmzYN48ePV0qQe/XqVS4q8kRERFS+KdhioaLUCfLSpUuxfv16lfE6deqgR48eOp8g79q1C5aWlsjPz0deXh46d+6MsWPHYsuWLWLi+5pcLoevr6/S2OsEGgAqVqyIWrVqITU1VRwzMjKCn5+f+H3t2rVha2uL1NRUNGrUCMnJyUhKSsKMGTPEOQUFBXj58iVevHgBc3NzAFBaozgymQwymUxpTBAESCT8i05ERET0rkqdIGdkZKBq1aoq45UrV0Z6enqZBKVOrVq1wuLFi2FsbAxHR0cYGxvjzz//BADs3r1bZScOqVT6j2u+mZAWlaC+HlMoFIiOjkbXrl1V5piamopfl2RXjdjYWERHRyuNGRhawcjI5h+fS0RERATwIr2ilDpBdnJywvHjx+Hq6qo0fvz4cTg6OpZZYOpiYWEBd3d3pTEvLy9IpVKkpaUptVMU5cSJE6hRowYAICsrC5cvX0bt2rXFx/Pz83Hq1Ck0atQIAHDp0iU8efJEnNOgQQNcunRJJYZ3ERERgdGjRyuNVars9a/XJSIiItJnpU6QBw4ciPDwcOTl5YnbvR08eBDjx4/HmDFjyjxATbCyssLYsWMxatQoKBQKNGvWDNnZ2UhMTISlpaXYLwwAU6dOhZ2dHezt7TFx4kRUqlQJwcHB4uPGxsYYPnw45s+fD2NjYwwbNgyNGzcWE+YpU6agQ4cOcHJyQrdu3WBgYICzZ8/i3Llzpd6tQiqVqlS42V5BREREpfF+bb9QNkqdII8fPx6PHz/GkCFDxDvqmZqaYsKECYiIiCjzADVl2rRpqFKlCmJjY3H9+nXY2tqiQYMG+Oabb5Tmffvttxg5ciSuXLmCevXqYceOHTAxMREfNzc3x4QJE9CrVy/cuXMHzZo1w6pVq8THg4KCsGvXLkydOhWzZs2CsbExateujYEDB2rstRIRERFR8SSCILzTLnXPnz9HamoqzMzM4OHhUaJe3fLs8OHDaNWqFbKysmBra1vknPj4eISHh+PJkycaja0wqamT1s5dlHf866UWulRdL3jPtkssSwY69OcE6Nbf4YaVPLQdgpJTD69oOwQqZ3Trp1u35Mnvau3c++x7qG3ttvc3qm1tdSp1Bfk1S0tLfPjhh2UZCxERERGR1pUoQe7atSvi4+NhbW1d5O4LhW3btq1MAiMiIiIi9eNnmqpKlCDb2NiIH0/b2OjnFmItW7b8x49aQ0NDERoaqpmAiIiIiMoAE2RVJUqQ4+LiivyaiIiIiOh98849yERERERU/vFGIapKlCD7+vqWeAeAv/76618FRERERESkTSVKkAvfCOPly5dYtGgRvLy8EBAQAODV3eUuXLiAIUOGqCVIIiIiIlIPBQvIKkqUIEdGRopfDxw4ECNGjMC0adNU5ty+fbtsoyMiIiIi0jCD0j7h559/Rt++fVXG+/Tpg61bt5ZJUERERESkGQpI1HaUV6VOkM3MzHDs2DGV8WPHjsHU1LRMgiIiIiIi0pZS72IRHh6Or776CsnJyWjcuDGAVz3Iq1atwpQpU8o8QCIiIiJSn7ff5UE/lTpB/vrrr+Hm5oYffvgB69evBwB4enoiPj4eISEhZR4gEREREakPbxSiSiL80+3hqFwxNqmm7RCU8C9X+WBQwm0cNaGkW0pqSoGC/3QUx6eii7ZDEJ17fFPbIYiMDAy1HYKSfEWBtkMQ6dZPN1DVsqK2QxClPT6ntXNvc+iltrW7ZqxX29rqVOoeZAB48uQJVqxYgW+++QaPHz8G8Gr/47t375ZpcERERESkXgqJRG1HeVXqFouzZ8+iTZs2sLGxwc2bNzFw4EBUrFgR27dvx61bt7BmzRp1xElEREREpBGlriCPHj0aoaGhuHLlitKuFe3bt8fvv/9epsERERERkXoJajzKq1InyElJSRg8eLDKeLVq1ZCRkVEmQRERERERaUupWyxMTU2RnZ2tMn7p0iVUrly5TIIiIiIiIs3gpciqSl1B7ty5M6ZOnYq8vDwAr644T0tLw9dff43PPvuszAMkIiIiItKkUifI33//PR48eIAqVaogNzcXLVq0gLu7O6ysrDBjxgx1xEhEREREaqKQqO8or0rdYmFtbY1jx47h0KFD+Ouvv6BQKNCgQQO0adNGHfERERERkRopdG6Hau0rVYKcn58PU1NTpKSk4OOPP8bHH3+srriIiIiIiLSiVC0WRkZGcHZ2RkGB7tyV598IDQ2FRCKBRCKBsbEx3NzcMHbsWOTk5Pzjcw8fPgyJRIInT56U6pw3b96ERCJBSkrKuwVNREREVIZ0ZZu3xYsXo27durC2toa1tTUCAgLw66+//i9OQUBUVBQcHR1hZmaGli1b4sKFC0pryGQyDB8+HJUqVYKFhQU6deqEO3fulDKSd+hBnjRpEiIiIsQ76JV37dq1Q3p6Oq5fv47p06dj0aJFGDt2rLbDIiIiItIr1atXx7fffotTp07h1KlT+Pjjj9G5c2cxCZ41axbmzJmDhQsXIikpCQ4ODggMDMSzZ8/ENcLDw7F9+3Zs3LgRx44dw/Pnz9GhQ4dSF3dLnSDPnz8fR48ehaOjI2rVqoUGDRooHeWNVCqFg4MDnJyc0KtXL/Tu3RsJCQmQyWQYMWIEqlSpAlNTUzRr1gxJSUkAXlWBW7VqBQCoUKECJBIJQkNDAQB79uxBs2bNYGtrCzs7O3To0AHXrl0Tz+fq6goA8PX1hUQiQcuWLQEACoUCU6dORfXq1SGVSlG/fn3s2bNHc28EERER6SVduUivY8eO+OSTT1CzZk3UrFkTM2bMgKWlJU6cOAFBEDBv3jxMnDgRXbt2hbe3N1avXo0XL15g/fr1AICnT59i5cqVmD17Ntq0aQNfX1+sW7cO586dw4EDB0oVS6kv0uvcuTMk5fje2v/EzMwMeXl5GD9+PLZu3YrVq1fD2dkZs2bNQlBQEK5evQonJyds3boVn332GS5dugRra2uYmZkBAHJycjB69Gj4+PggJycHU6ZMQZcuXZCSkgIDAwOcPHkSjRo1woEDB1CnTh2YmJgAAH744QfMnj0bS5cuha+vL1atWoVOnTrhwoUL8PDw0OZbQkRERPROZDIZZDKZ0phUKoVUKn3r8woKCvDzzz8jJycHAQEBuHHjBjIyMtC2bVuldVq0aIHExEQMHjwYycnJyMvLU5rj6OgIb29vJCYmIigoqMRxlzpBjoqKKu1Tyo2TJ09i/fr1aNWqFRYvXoz4+Hi0b98eALB8+XLs378fK1euxLhx41CxYkUAQJUqVWBrayuu8eZe0CtXrkSVKlVw8eJFeHt7izdTsbOzg4ODgzjv+++/x4QJE9CjRw8AwMyZM/Hbb79h3rx5+PHHH9X5somIiEiPqfNGIbGxsYiOjlYai4yMLDafPHfuHAICAvDy5UtYWlpi+/bt8PLyQmJiIgDA3t5eab69vT1u3boFAMjIyICJiQkqVKigMqe0d3sucYvFixcvMHToUFSrVg1VqlRBr1698PDhw1KdTBft2rULlpaWMDU1RUBAAJo3b47hw4cjLy8PTZs2FecZGxujUaNGSE1Nfet6165dQ69eveDm5gZra2uxpSItLa3Y52RnZ+PevXtK5wOApk2bvvV8MpkM2dnZSocglOc7nxMREdH7JCIiAk+fPlU6IiIiip1fq1YtpKSk4MSJE/jqq6/Qr18/XLx4UXz8zS4GQRD+sbOhJHPeVOIEOTIyEvHx8fj000/Ro0cP7N+/H1999VWpTqaLWrVqhZSUFFy6dAkvX77Etm3bYGNjA+Dd/hA6duyIR48eYfny5fjzzz/x559/AgDkcvk/xlLa88XGxsLGxkbpUCieFTufiIiI6E3q3MVCKpWKu1K8Pt7WXmFiYgJ3d3f4+fkhNjYW9erVww8//CB+6v5mJTgzM1OsKjs4OEAulyMrK6vYOSVV4gR527ZtWLlyJZYtW4b58+dj9+7dSEhIKPdbvllYWMDd3R3Ozs4wNjYGALi7u8PExATHjh0T5+Xl5eHUqVPw9PQEALF3uPDrf/ToEVJTUzFp0iS0bt0anp6eKn9IRT3P2toajo6OSucDgMTERPF8RSnqtzIDA6t3eRuIiIhIT+nKRXpFEQQBMpkMrq6ucHBwwP79+8XH5HI5jhw5giZNmgAAGjZsCGNjY6U56enpOH/+vDinpErcg3z79m189NFH4veNGjWCkZER7t27Bycnp1KdVNdZWFjgq6++EnuNa9SogVmzZuHFixcICwsDADg7O0MikWDXrl345JNPYGZmhgoVKsDOzg7Lli1D1apVkZaWhq+//lpp7SpVqsDMzAx79uxB9erVYWpqChsbG4wbNw6RkZH44IMPUL9+fcTFxSElJQU//fRTsXEW1eT+Pl9ASURERO+vb775Bu3bt4eTkxOePXuGjRs34vDhw9izZw8kEgnCw8MRExMDDw8PeHh4ICYmBubm5ujVqxcAwMbGBmFhYRgzZgzs7OxQsWJFjB07Fj4+PqW+43OJE+SCggKx+ik+2cgI+fn5pTphefHtt99CoVDg888/x7Nnz+Dn54e9e/eKjd/VqlVDdHQ0vv76a/Tv3x99+/ZFfHw8Nm7ciBEjRsDb2xu1atXC/Pnzxa3cgFfv2fz58zF16lRMmTIFH330EQ4fPowRI0YgOzsbY8aMQWZmJry8vLBjxw7uYEFERERqpc6L9Erj/v37+Pzzz5Geng4bGxvUrVsXe/bsQWBgIABg/PjxyM3NxZAhQ5CVlQV/f3/s27cPVlb/+/R87ty5MDIyQkhICHJzc9G6dWvEx8fD0NCwVLFIhBJe1WVgYID27dsrVSx37tyJjz/+GBYWFuLYtm3bShUAlS1jk2raDkEJLxksHwx06JMHXfsUpEChK/906B6fii7aDkF07vFNbYcgMjIo3T/E6pav0J1WSN366QaqWlbUdgiitMfntHbu5dX7qG3tQXfWqW1tdSpxBblfv34qY336qO8NJSIiIiL1YxlAVYkT5Li4OHXGQURERESkE0p9oxAiIiIien8Iutb7ogNKvM0bEREREZE+YAWZiIiISI+xB1kVE2QiIiIiPcYEWRVbLIiIiIiICmEFmYiIiEiP8Z4FqlhBJiIiIiIqhBVkIiIiIj2m4DZvKlhBJiIiIiIqhBVkIiIiIj3GXSxUMUF+z7DRnt6FQtChvzm6FAu91bnHN7Udgshaaq7tEETZshfaDkFn6dpP973nj7UdAukoJshEREREeowVZFVMkImIiIj0mK5V9nUBL9IjIiIiIiqEFWQiIiIiPcZt3lSxgkxEREREVAgryERERER6jBfpqWIFmYiIiIioEFaQiYiIiPQYd7FQxQpyMUJDQxEcHKztMIiIiIhIw3QyQQ4NDYVEIoFEIoGRkRFq1KiBr776CllZWdoOjYiIiOi9ooCgtqO80skEGQDatWuH9PR03Lx5EytWrMDOnTsxZMgQbYf1rwiCgPz8fJVxuVyuhWiIiIiIXl2kp66jvNLZBFkqlcLBwQHVq1dH27Zt0b17d+zbtw8AUFBQgLCwMLi6usLMzAy1atXCDz/8oPT8w4cPo1GjRrCwsICtrS2aNm2KW7duAQCioqJQv359LF26FE5OTjA3N0e3bt3w5MkTlTiio6NRpUoVWFtbY/DgwUrJrCAImDVrFtzc3GBmZoZ69ephy5YtSjFIJBLs3bsXfn5+kEqlOHr0KFq2bIlhw4Zh9OjRqFSpEgIDAzFgwAB06NBB6dz5+flwcHDAqlWryuptJSIiIqJ/UC4u0rt+/Tr27NkDY2NjAIBCoUD16tWxefNmVKpUCYmJifjiiy9QtWpVhISEID8/H8HBwRg0aBA2bNgAuVyOkydPQiL5307YV69exebNm7Fz505kZ2cjLCwMQ4cOxU8//STOOXjwIExNTfHbb7/h5s2b6N+/PypVqoQZM2YAACZNmoRt27Zh8eLF8PDwwO+//44+ffqgcuXKaNGihbjO+PHj8f3338PNzQ22trYAgNWrV+Orr77C8ePHIQgCHj9+jObNmyM9PR1Vq1YFAPzyyy94/vw5QkJC1P0WExERkZ4qv40Q6qOzCfKuXbtgaWmJgoICvHz5EgAwZ84cAICxsTGio6PFua6urkhMTMTmzZsREhKC7OxsPH36FB06dMAHH3wAAPD09FRa/+XLl1i9ejWqV68OAFiwYAE+/fRTzJ49Gw4ODgAAExMTrFq1Cubm5qhTpw6mTp2KcePGYdq0acjNzcWcOXNw6NAhBAQEAADc3Nxw7NgxLF26VClBnjp1KgIDA5XO7+7ujlmzZimN1apVC2vXrsX48eMBAHFxcejWrRssLS3/3ZtJRERERCWmswlyq1atsHjxYrx48QIrVqzA5cuXMXz4cPHxJUuWYMWKFbh16xZyc3Mhl8tRv359AEDFihURGhqKoKAgBAYGok2bNggJCRErswBQo0YNMTkGgICAACgUCly6dElMkOvVqwdzc3OlOc+fP8ft27eRmZmJly9fqiS+crkcvr6+SmN+fn4qr6+osYEDB2LZsmUYP348MjMzsXv3bhw8eLDY90gmk0EmkymNCYKgVCknIiIiepvy3CusLjrbg2xhYQF3d3fUrVsX8+fPh0wmE6vGmzdvxqhRozBgwADs27cPKSkp6N+/v1J/cFxcHP744w80adIEmzZtQs2aNXHixIliz/c6qSxJcimRSKBQvPrrtHv3bqSkpIjHxYsXlfqQX7+Wol7fm/r27Yvr16/jjz/+wLp16+Di4oKPPvqo2DhiY2NhY2OjdAiKZ/8YPxEREREVT2cryG+KjIxE+/bt8dVXX+Ho0aNo0qSJ0q4W165dU3mOr68vfH19ERERgYCAAKxfvx6NGzcGAKSlpeHevXtwdHQEAPzxxx8wMDBAzZo1xeefOXMGubm5MDMzAwCcOHEClpaWqF69OipUqACpVIq0tDSldop/w87ODsHBwWJy379//7fOj4iIwOjRo5XGKtjVLpNYiIiISD8o+MGzinKTILds2RJ16tRBTEwMPDw8sGbNGuzduxeurq5Yu3YtkpKS4OrqCgC4ceMGli1bhk6dOsHR0RGXLl3C5cuX0bdvX3E9U1NT9OvXD99//z2ys7MxYsQIhISEiO0VwKt2ibCwMEyaNAm3bt1CZGQkhg0bBgMDA1hZWWHs2LEYNWoUFAoFmjVrhuzsbCQmJsLS0hL9+vV7p9c5cOBAdOjQAQUFBf+4hlQqhVQqVRpjewURERHRv1NuEmQAGD16NPr374/Lly8jJSUF3bt3h0QiQc+ePTFkyBD8+uuvAABzc3P8/fffWL16NR49eoSqVati2LBhGDx4sLiWu7s7unbtik8++QSPHz/GJ598gkWLFimdr3Xr1vDw8EDz5s0hk8nQo0cPREVFiY9PmzYNVapUQWxsLK5fvw5bW1s0aNAA33zzzTu/xjZt2qBq1aqoU6eOWN0mIiIiUpfyfEMPdZEIgqB370pUVBQSEhKQkpKi7VBUvHjxAo6Ojli1ahW6du1a6ucbmVRTQ1RERLrNWmr+z5M0JFv2QtshUDmUL7+rtXNPdOmltrVn3FyvtrXVqVxVkN9nCoUCGRkZmD17NmxsbNCpUydth0RERESkl5gg64i0tDS4urqievXqiI+Ph5ER/2iIiIhI/bjNmyq9bLF4n7HFgoj0EVssqLzTZotFhBpbLGLZYkFERERE5Q0v0lOlszcKISIiIiLSBlaQiYiIiPQY68eqWEEmIiIiIiqEFWQiIiIiPcZdLFQxQSYiIiLSY7xITxVbLIiIiIiICmEFmYiIiEiPsX6sihVkIiIiIqJCWEF+z0i0HcAbdOm3Ul16b3TpfSF6H+jS3euGODbTdghKFt07pu0QSMfxIj1VrCATERERERXCCjIRERGRHhP4uaYKVpCJiIiIiAphBZmIiIhIj7EHWRUryERERESkdbGxsfjwww9hZWWFKlWqIDg4GJcuXVKaIwgCoqKi4OjoCDMzM7Rs2RIXLlxQmiOTyTB8+HBUqlQJFhYW6NSpE+7cuVOqWJggExEREekxBQS1HaVx5MgRDB06FCdOnMD+/fuRn5+Ptm3bIicnR5wza9YszJkzBwsXLkRSUhIcHBwQGBiIZ8+eiXPCw8Oxfft2bNy4EceOHcPz58/RoUMHFBQUlDgWiSAI7Mx+jxibVNN2CEp06S8Xt3kjIk3gNm/0LvLld7V27q9cQtS29uKbm9/5uQ8ePECVKlVw5MgRNG/eHIIgwNHREeHh4ZgwYQKAV9Vie3t7zJw5E4MHD8bTp09RuXJlrF27Ft27dwcA3Lt3D05OTvjll18QFBRUonOzgkxEREREaiGTyZCdna10yGSyEj336dOnAICKFSsCAG7cuIGMjAy0bdtWnCOVStGiRQskJiYCAJKTk5GXl6c0x9HREd7e3uKckmCC/A8OHz4MiUSCJ0+eaDsUIiIiojKnzhaL2NhY2NjYKB2xsbH/GJMgCBg9ejSaNWsGb29vAEBGRgYAwN7eXmmuvb29+FhGRgZMTExQoUKFYueURLlPkDMzMzF48GDUqFEDUqkUDg4OCAoKwh9//KHt0IiIiIj0WkREBJ4+fap0RERE/OPzhg0bhrNnz2LDhg0qj0kkyk2TgiCojL2pJHMKK/fbvH322WfIy8vD6tWr4ebmhvv37+PgwYN4/PixtkMjIiIi0nnq3OZNKpVCKpWW6jnDhw/Hjh078Pvvv6N69eriuIODA4BXVeKqVauK45mZmWJV2cHBAXK5HFlZWUpV5MzMTDRp0qTEMZTrCvKTJ09w7NgxzJw5E61atYKzszMaNWqEiIgIfPrppwCAtLQ0dO7cGZaWlrC2tkZISAju378PALh58yYMDAxw6tQppXUXLFgAZ2dnFL5+8fjx46hXrx5MTU3h7++Pc+fOKT0nMTERzZs3h5mZGZycnDBixAilqy7XrVsHPz8/WFlZwcHBAb169UJmZqb4+OtWjoMHD8LPzw/m5uZo0qSJyvYmRERERO8jQRAwbNgwbNu2DYcOHYKrq6vS466urnBwcMD+/fvFMblcjiNHjojJb8OGDWFsbKw0Jz09HefPn9efBNnS0hKWlpZISEgosuFbEAQEBwfj8ePHOHLkCPbv349r166JVzW6uLigTZs2iIuLU3peXFwcQkNDlUrx48aNw/fff4+kpCRUqVIFnTp1Ql5eHgDg3LlzCAoKQteuXXH27Fls2rQJx44dw7Bhw8Tny+VyTJs2DWfOnEFCQgJu3LiB0NBQlZgnTpyI2bNn49SpUzAyMsKAAQPK4q0iIiIiKpKgxv9KY+jQoVi3bh3Wr18PKysrZGRkICMjA7m5uQBetVaEh4cjJiYG27dvx/nz5xEaGgpzc3P06tULAGBjY4OwsDCMGTMGBw8exOnTp9GnTx/4+PigTZs2JY6l3G/ztnXrVgwaNAi5ublo0KABWrRogR49eqBu3brYv38/2rdvjxs3bsDJyQkAcPHiRdSpUwcnT57Ehx9+iM2bN+PLL79Eeno6pFIpzpw5A19fX1y/fh0uLi44fPgwWrVqhY0bN4qJ9ePHj1G9enXEx8cjJCQEffv2hZmZGZYuXSrGdezYMbRo0QI5OTkwNTVViTspKQmNGjXCs2fPYGlpKZ7nwIEDaN26NQDgl19+waefforc3Nwi1ygKt3krHrd5IyJN4DZv9C60uc3bQJf/qG3tFTe3lHhucT3CrwuXwKviZ3R0NJYuXYqsrCz4+/vjxx9/FC/kA4CXL19i3LhxWL9+PXJzc9G6dWssWrRIzAVLolxXkIFXPcj37t3Djh07EBQUhMOHD6NBgwaIj49HamoqnJyclN4QLy8v2NraIjU1FQAQHBwMIyMjbN++HQCwatUqtGrVCi4uLkrnCQgIEL+uWLEiatWqJa6RnJyM+Ph4saJtaWmJoKAgKBQK3LhxAwBw+vRpdO7cGc7OzrCyskLLli0BvGoBKaxu3bri16/7awq3YhRW1NYp5fz3HSIiItIwhRqP0hAEocij8CfuEokEUVFRSE9Px8uXL3HkyBGl5BgATE1NsWDBAjx69AgvXrzAzp07S5UcA+9Bggy8eiMCAwMxZcoUJCYmIjQ0FJGRkcVesVh43MTEBJ9//jni4uIgl8uxfv36Erc1vF5DoVBg8ODBSElJEY8zZ87gypUr+OCDD5CTk4O2bdvC0tIS69atQ1JSkpiQy+VypTWNjY2LXL8oRW2dolA8K3IuERERUVF0pcVCl5T7XSyK4uXlhYSEBHh5eSEtLQ23b99WarF4+vQpPD09xfkDBw6Et7c3Fi1ahLy8PHTt2lVlzRMnTqBGjRoAgKysLFy+fBm1a9cGADRo0AAXLlyAu7t7kfGcO3cODx8+xLfffivG8eaFge8iIiICo0ePVhqraFf7X69LREREpM/KdYL86NEjdOvWDQMGDEDdunVhZWWFU6dOYdasWejcuTPatGmDunXronfv3pg3bx7y8/MxZMgQtGjRAn5+fuI6np6eaNy4MSZMmIABAwbAzMxM5VxTp06FnZ0d7O3tMXHiRFSqVAnBwcEAgAkTJqBx48YYOnQoBg0aBAsLC6SmpmL//v1YsGABatSoARMTEyxYsABffvklzp8/j2nTpv3r11/U1iml2eOPiIiISJ3bvJVX5brFwtLSEv7+/pg7dy6aN28Ob29vTJ48GYMGDcLChQshkUiQkJCAChUqoHnz5mjTpg3c3NywadMmlbXCwsIgl8uLba/49ttvMXLkSDRs2BDp6enYsWMHTExMALzqGz5y5AiuXLmCjz76CL6+vpg8ebLYQ1y5cmXEx8fj559/hpeXF7799lt8//336ntjiIiIiOidlftdLMrKjBkzsHHjRpX9jcsb7mJRPF2qrevS+0JEZYu7WNC70OYuFp87q7aWlpW1t7apbW11KtcV5LLw/PlzJCUlYcGCBRgxYoS2wyEiIiIiLdP7BHnYsGFo1qwZWrRowZtyEBERkd4R1HiUV+X6Ir2yEB8fj/j4eG2HQUREREQ6Qu8TZCIiIiJ9pijXtV71YIJMREREpMfK8w091EXve5CJiIiIiApjBZmIiIhIj/FGIapYQSYiIiIiKoQVZCIiIiI9xov0VLGCTERERERUCCvIRERERHqMu1ioYgWZiIiIiKgQVpDfM/wdsHh8b4hIExbdO6btEJS42zpqOwTR1Sf3tB0CFYG7WKhigkxERESkxwSBJaQ3scWCiIiIiKgQVpCJiIiI9Bi3eVPFCjIRERERUSGsIBMRERHpMV6kp4oVZCIiIiKiQlhBJiIiItJjvFGIKlaQiYiIiIgKYQVZR4SGhuLJkydISEjQdihERESkR7iLhSq9qCBnZmZi8ODBqFGjBqRSKRwcHBAUFIQ//vhD26ERERERaZUgCGo7yiu9qCB/9tlnyMvLw+rVq+Hm5ob79+/j4MGDePz4sbZDIyIiIiId895XkJ88eYJjx45h5syZaNWqFZydndGoUSNERETg008/BQBcuXIFzZs3h6mpKby8vLB//35IJBKx3eHw4cOQSCR48uSJuG5KSgokEglu3rwJAHj06BF69uyJ6tWrw9zcHD4+PtiwYYNSLFu2bIGPjw/MzMxgZ2eHNm3aICcnR2nO999/j6pVq8LOzg5Dhw5FXl6e2t4bIiIiIoUaj/Lqva8gW1pawtLSEgkJCWjcuDGkUqnS4wqFAl27dkWlSpVw4sQJZGdnIzw8vNTnefnyJRo2bIgJEybA2toau3fvxueffw43Nzf4+/sjPT0dPXv2xKxZs9ClSxc8e/YMR48eVfr44bfffkPVqlXx22+/4erVq+jevTvq16+PQYMG/du3gYiIiIhK6L1PkI2MjBAfH49BgwZhyZIlaNCgAVq0aIEePXqgbt26OHDgAFJTU3Hz5k1Ur14dABATE4P27duX6jzVqlXD2LFjxe+HDx+OPXv24OeffxYT5Pz8fHTt2hXOzs4AAB8fH6U1KlSogIULF8LQ0BC1a9fGp59+ioMHDzJBJiIiIrXhNm+q3vsWC+BVD/K9e/ewY8cOBAUF4fDhw2jQoAHi4+ORmpqKGjVqiMkxAAQEBJT6HAUFBZgxYwbq1q0LOzs7WFpaYt++fUhLSwMA1KtXD61bt4aPjw+6deuG5cuXIysrS2mNOnXqwNDQUPy+atWqyMzMLPacMpkM2dnZSkd5bognIiIi0gV6kSADgKmpKQIDAzFlyhQkJiYiNDQUkZGRRSaUEolE6XsDg1dvU+G5b/YGz549G3PnzsX48eNx6NAhpKSkICgoCHK5HABgaGiI/fv349dff4WXlxcWLFiAWrVq4caNG+IaxsbGKnEoFMV38MTGxsLGxkbpEBTPSviOEBEREb3a5k1dR3mlNwnym7y8vJCTkwMvLy+kpaXh3r174mNvbv9WuXJlAEB6ero4lpKSojTn6NGj6Ny5M/r06YN69erBzc0NV65cUZojkUjQtGlTREdH4/Tp0zAxMcH27dvf+TVERETg6dOnSofEwOqd1yMiIiIiPehBfvToEbp164YBAwagbt26sLKywqlTpzBr1ix07twZbdq0Qa1atdC3b1/Mnj0b2dnZmDhxotIa7u7ucHJyQlRUFKZPn44rV65g9uzZKnO2bt2KxMREVKhQAXPmzEFGRgY8PT0BAH/++ScOHjyItm3bokqVKvjzzz/x4MED8fF3IZVKVS46fLP6TURERPQ2bM9U9d4nyJaWlvD398fcuXNx7do15OXlwcnJCYMGDcI333wDAwMDbN++HWFhYWjUqBFcXFwwf/58tGvXTlzD2NgYGzZswFdffYV69erhww8/xPTp09GtWzdxzuTJk3Hjxg0EBQXB3NwcX3zxBYKDg/H06VMAgLW1NX7//XfMmzcP2dnZcHZ2xuzZs0t9MSARERFRWSrPrRDqIhH4a0ORJBIJtm/fjuDgYG2HUipGJtW0HQIREekQd1tHbYcguvrk3j9P0lP58rtaO3er6oFqW/u3O/vVtrY6vfcVZCIiIiIqHrd5U6W3F+kRERERERWFFeRisPOEiIiI9IGCOY8KVpCJiIiIiAphBZmIiIhIj7F+rIoVZCIiIiKiQlhBJiIiItJj3AdZFRNkIiIiIj3GBFkVWyyIiIiIiAphBZmIiIhIj3FrW1WsIBMRERERFcIKMhEREZEeYw+yKibIRFpgaKBbH97o0sdrvKNT+WEgkWg7BBH/3hTv6pN72g5B5G7rqO0QlFx/mq7tEOgNv//+O7777jskJycjPT0d27dvR3BwsPi4IAiIjo7GsmXLkJWVBX9/f/z444+oU6eOOEcmk2Hs2LHYsGEDcnNz0bp1ayxatAjVq1cvcRy69a80EREREWmUoMb/SisnJwf16tXDwoULi3x81qxZmDNnDhYuXIikpCQ4ODggMDAQz549E+eEh4dj+/bt2LhxI44dO4bnz5+jQ4cOKCgoKHEcEkGXSkf0rxmZVNN2CFQCrCAXj5XA8oMVZCotVpCLJ5fd0dq5P3Rsrra1k+79/s7PlUgkShVkQRDg6OiI8PBwTJgwAcCrarG9vT1mzpyJwYMH4+nTp6hcuTLWrl2L7t27AwDu3bsHJycn/PLLLwgKCirRuXXrX2kiIiIi0ihBENR2lKUbN24gIyMDbdu2FcekUilatGiBxMREAEBycjLy8vKU5jg6OsLb21ucUxLsQSYiIiLSY+q8SE8mk0EmkymNSaVSSKXSUq+VkZEBALC3t1cat7e3x61bt8Q5JiYmqFChgsqc188vCVaQiYiIiEgtYmNjYWNjo3TExsb+qzUlb7R3CYKgMvamkswpjBVkIiIiIj2mzutQIiIiMHr0aKWxd6keA4CDgwOAV1XiqlWriuOZmZliVdnBwQFyuRxZWVlKVeTMzEw0adKkxOdiBZmIiIiI1EIqlcLa2lrpeNcE2dXVFQ4ODti/f784JpfLceTIETH5bdiwIYyNjZXmpKen4/z586VKkFlBJiIiItJjunSjkOfPn+Pq1avi9zdu3EBKSgoqVqyIGjVqIDw8HDExMfDw8ICHhwdiYmJgbm6OXr16AQBsbGwQFhaGMWPGwM7ODhUrVsTYsWPh4+ODNm3alDgOJsjFaNmyJerXr4958+ZpOxQiIiIivXDq1Cm0atVK/P51e0a/fv0QHx+P8ePHIzc3F0OGDBFvFLJv3z5YWVmJz5k7dy6MjIwQEhIi3igkPj4ehoaGJY5Dq/sgZ2ZmYvLkyfj1119x//59VKhQAfXq1UNUVBQCAgJU9r/TpPKaIHMf5PKB+yAXj/vZlh/cB5lKi/sgF0+b+yDXdQhQ29pnM/5Q29rqpNUK8meffYa8vDysXr0abm5uuH//Pg4ePIjHjx+XeI28vDwYGxurMcqyU1BQAIlEAoM3kiO5XA4TExMtRUVEREREhWmtjPXkyRMcO3YMM2fORKtWreDs7IxGjRohIiICn376KVxcXAAAXbp0gUQiEb+PiopC/fr1sWrVKri5uUEqlUIQBOzZswfNmjWDra0t7Ozs0KFDB1y7dk0832effYbhw4eL34eHh0MikeDChQsAgPz8fFhZWWHv3r3inPz8fAwbNkxcc9KkSUqVNrlcjvHjx6NatWqwsLCAv78/Dh8+LD4eHx8PW1tb7Nq1C15eXpBKpbh16xZcXFwwffp0hIaGwsbGBoMGDcLHH3+MYcOGKb1Hjx49glQqxaFDh8rqbSciIiJSohAEtR3lldYSZEtLS1haWiIhIUFlA2kASEpKAgDExcUhPT1d/B4Arl69is2bN2Pr1q1ISUkB8Ore3aNHj0ZSUhIOHjwIAwMDdOnSBQqFAsCrlonCyeuRI0dQqVIlHDlyRDzfy5cv0bRpU3HO6tWrYWRkhD///BPz58/H3LlzsWLFCvHx/v374/jx49i4cSPOnj2Lbt26oV27drhy5Yo458WLF4iNjcWKFStw4cIFVKlSBQDw3XffwdvbG8nJyZg8eTIGDhyI9evXK70XP/30ExwdHZV6cYiIiIhIvbTag7x161YMGjQIubm5aNCgAVq0aIEePXqgbt26r4Irogc5KioKMTExuHv3LipXrlzs2g8ePECVKlVw7tw5eHt749y5c6hXrx4yMzNhaGgIe3t7REZG4syZM9i8eTNiY2Px3//+FydOnADwKqHOzMzEhQsXxI2lv/76a+zYsQMXL17EtWvX4OHhgTt37sDR8X89VW3atEGjRo0QExOD+Ph49O/fHykpKahXr544x8XFBb6+vti+fbs4JpPJ4OjoiMWLFyMkJAQA4Ovri+DgYERGRpb4PWUPcvnAHuTileeKg75hDzKVFnuQi6fNHuQ69v5qW/vC/T/VtrY6afVf6c8++wz37t3Djh07EBQUhMOHD6NBgwaIj49/6/OcnZ1VkuNr166hV69ecHNzg7W1NVxdXQEAaWlpAABvb2/Y2dnhyJEjOHr0KOrVq4dOnTqJFeTDhw+jRYsWSms2btxY6a4rAQEBuHLlCgoKCvDXX39BEATUrFlTrIZbWlriyJEjSq0dJiYmYsJfmJ+fn9L3UqkUffr0wapVqwAAKSkpOHPmDEJDQ4t9H2QyGbKzs5UOXUp0iIiISPexxUKV1rd5MzU1RWBgIAIDAzFlyhQMHDgQkZGRb00MLSwsVMY6duwIJycnLF++HI6OjlAoFPD29oZcLgfwqhrdvHlzHD58GCYmJmjZsiW8vb1RUFCAc+fOITExEeHh4SWOW6FQwNDQEMnJySrbhlhaWopfm5mZFXlrw6Jew8CBA1G/fn3cuXMHq1atQuvWreHs7FxsDLGxsYiOjlYakxhYQmJoXeLXQURERETKdOtzXgBeXl7IyckBABgbG6OgoOAfn/Po0SOkpqZi0qRJaN26NTw9PZGVlaUy73Uf8uHDh9GyZUtIJBJ89NFH+P7775Gbm6vUfwxAbLco/L2HhwcMDQ3h6+uLgoICZGZmwt3dXel4fSvE0vLx8YGfnx+WL1+O9evXY8CAAW+dHxERgadPnyodEgOrtz6HiIiIqDBBjf+VV1pLkB89eoSPP/4Y69atw9mzZ3Hjxg38/PPPmDVrFjp37gzgVa/uwYMHkZGRUWTC+1qFChVgZ2eHZcuW4erVqzh06JDKfb+BVwnyhQsXcO7cOXz00Ufi2E8//YQGDRrA2lq58nr79m2MHj0aly5dwoYNG7BgwQKMHDkSAFCzZk307t0bffv2xbZt23Djxg0kJSVh5syZ+OWXX975fRk4cCC+/fZbFBQUoEuXLm+dW9TtG4uqVhMRERFRyWmtxcLS0hL+/v6YO3curl27hry8PDg5OWHQoEH45ptvAACzZ8/G6NGjsXz5clSrVg03b94sci0DAwNs3LgRI0aMgLe3N2rVqoX58+ejZcuWSvO8vb1RqVIlODs7i8lwixYtUFBQoNJ/DAB9+/ZFbm4uGjVqBENDQwwfPhxffPGF+HhcXBymT5+OMWPG4O7du7Czs0NAQAA++eSTd35fevbsifDwcPTq1QumpqbvvA4RERFRSZTnXmF10eouFqTq9u3bcHFxQVJSEho0aFDq53MXi/KBu1gUj/+jLj+4iwWVFnexKJ42d7GoWdnvnye9o8sPTqltbXXS+kV69EpeXh7S09Px9ddfo3Hjxu+UHBMRERGVVnnuFVYX3Spj6bHjx4/D2dkZycnJWLJkibbDISIiItJbrCDriJYtW+rUx9xERESkH9iipIoJMhEREZEeY4uFKrZYEBEREREVwgoyERERkR4TBIW2Q9A5rCATERERERXCCjIRERGRHlOwB1kFK8hERERERIWwgkxERESkx7jNrCpWkImIiIiICmEFmYiIiEiPsQdZFRPk94yhgW59KFCg0J2tYyTaDqAQXXpfdI2BRJf+pHiHqbfhx7JFMzIw1HYISvIVBdoOQXT1yT1th6Ak995RbYegE/izrEq3sikiIiIiIi1jBZmIiIhIj/GTMlWsIBMRERERFcIKMhEREZEeE3iRngpWkImIiIiICmEFmYiIiEiPcRcLVawgExEREREVwgoyERERkR7jjUJUMUEmIiIi0mNssVD13rdYSCSStx6hoaFaicvFxQXz5s3TyrmJiIiIqHjvfQU5PT1d/HrTpk2YMmUKLl26JI6ZmZlpIywiIiIincAbhah67yvIDg4O4mFjYwOJRKI0tn79enzwwQcwMTFBrVq1sHbtWqXnSyQSLF26FB06dIC5uTk8PT3xxx9/4OrVq2jZsiUsLCwQEBCAa9euic+5du0aOnfuDHt7e1haWuLDDz/EgQMHxMdbtmyJW7duYdSoUWIl+7WtW7eiTp06kEqlcHFxwezZs9X/JhERERGR6L1PkN9m+/btGDlyJMaMGYPz589j8ODB6N+/P3777TeledOmTUPfvn2RkpKC2rVro1evXhg8eDAiIiJw6tQpAMCwYcPE+c+fP8cnn3yCAwcO4PTp0wgKCkLHjh2RlpYGANi2bRuqV6+OqVOnIj09XaxyJycnIyQkBD169MC5c+cQFRWFyZMnIz4+XjNvCBEREekdQRDUdpRXEqE8R19K8fHxCA8Px5MnTwAATZs2RZ06dbBs2TJxTkhICHJycrB7924AryrIkyZNwrRp0wAAJ06cQEBAAFauXIkBAwYAADZu3Ij+/fsjNze32HPXqVMHX331lZhIu7i4IDw8HOHh4eKc3r1748GDB9i3b584Nn78eOzevRsXLlwo0WuUmjqVaJ6mFCgU2g5BJPnnKRqjNz9078BAokt/Uvzo8W106U9Kl/6UjAwMtR2CknxFgbZD0Fm5945qOwSRcSU3rZ27gqW72tbOen5VbWurk15XkFNTU9G0aVOlsaZNmyI1NVVprG7duuLX9vb2AAAfHx+lsZcvXyI7OxsAkJOTg/Hjx8PLywu2trawtLTE33//LVaQSxvPlStXUFCg+j84mUyG7OxspUOPft8hIiKiMqCAoLajvNLrBBmAUv8v8OpjhjfHjI2NVeYXNab4/2rpuHHjsHXrVsyYMQNHjx5FSkoKfHx8IJfL3xpLUed+W8IbGxsLGxsbpaOgIPut5yAiIiKit9PrBNnT0xPHjh1TGktMTISnp+e/Wvfo0aMIDQ1Fly5d4OPjAwcHB9y8eVNpjomJiUpV2MvLq8h4atasCUND1Y/sIiIi8PTpU6XD0ND6X8VORERE+oU9yKre+23e3mbcuHEICQlBgwYN0Lp1a+zcuRPbtm1T2nHiXbi7u2Pbtm3o2LEjJBIJJk+eLFaXX3NxccHvv/+OHj16QCqVolKlShgzZgw+/PBDTJs2Dd27d8cff/yBhQsXYtGiRUWeRyqVQiqVKo29WYEmIiIiehtea6FKryvIwcHB+OGHH/Ddd9+hTp06WLp0KeLi4tCyZct/te7cuXNRoUIFNGnSBB07dkRQUBAaNGigNGfq1Km4efMmPvjgA1SuXBkA0KBBA2zevBkbN26Et7c3pkyZgqlTp2rtZiZERERE+kivdrHQB9zFoni6VFvnD13xuItF+aFLf1K69KfEXSzKD+5i8YqFuYva1s55cVNta6uTXleQiYiIiIjepNc9yERERET6jp+UqWIFmYiIiIioEFaQiYiIiPQYL0dTxQoyEREREVEhrCATERER6TFBp/aB0Q1MkImIiIj0GFssVLHFgoiIiIh0xqJFi+Dq6gpTU1M0bNgQR49qfr9qJshEREREekwQBLUdpbVp0yaEh4dj4sSJOH36ND766CO0b98eaWlpanjlxeOd9N4zvJNe8XjXr/KBd9IrP3TpT0qX/pR4J73yg3fS+/9zm1RT29p58rulmu/v748GDRpg8eLF4pinpyeCg4MRGxtb1uEVixVkIiIiIj0mqPEoDblcjuTkZLRt21ZpvG3btkhMTHyXl/bOeJEeEREREamFTCaDTCZTGpNKpZBKpSpzHz58iIKCAtjb2yuN29vbIyMjQ61xqhCI3vDy5UshMjJSePnypbZD0alYBEG34tGlWARBt+JhLMXTpXh0KRZB0K14dCkWQdCteBhL+RIZGalSWI6MjCxy7t27dwUAQmJiotL49OnThVq1amkg2v9hDzKpyM7Oho2NDZ4+fQpra2vGoqPx6FIsuhYPYykf8ehSLLoWjy7FomvxMJbypTQVZLlcDnNzc/z888/o0qWLOD5y5EikpKTgyJEjao/3NfYgExEREZFaSKVSWFtbKx1FJccAYGJigoYNG2L//v1K4/v370eTJk00Ea6IPchEREREpBNGjx6Nzz//HH5+fggICMCyZcuQlpaGL7/8UqNxMEEmIiIiIp3QvXt3PHr0CFOnTkV6ejq8vb3xyy+/wNnZWaNxMEEmFVKpFJGRkcV+BKKvsQC6FY8uxQLoVjyMpXi6FI8uxQLoVjy6FAugW/EwlvffkCFDMGTIEK3GwIv0iIiIiIgK4UV6RERERESFMEEmIiIiIiqECTIRERERUSFMkImIiIiICmGCTERERFRKgiCA+xy8v5ggEwBg7dq1aNq0KRwdHXHr1i0AwLx58/Df//5XK/Hk5ORo5bxvcnFxwdSpU5GWlqbtUERPnjzBvn37sG7dOqxZs0bp0LSjR4+iT58+CAgIwN27dwG8+rt07NgxjccCANeuXcOkSZPQs2dPZGZmAgD27NmDCxcuaCUeXZCfn4/o6Gjcvn1b26FQCcnlcty5cwdpaWlKhz47fPiwtkMQrVmzBj4+PjAzM4OZmRnq1q2LtWvXajssKmsC6b1FixYJlSpVEqZPny6YmZkJ165dEwRBEOLi4oSWLVtqJSYLCwuhf//+wtGjR7Vy/tfmz58vNGjQQDA0NBTatGkjbNiwQXj58qXW4tmxY4dgZWUlGBgYCDY2NoKtra14VKhQQaOxbNmyRTAzMxMGDhwoSKVS8e/Njz/+KLRv316jsQiCIBw+fFgwMzMT2rRpI5iYmIjxzJw5U/jss880Ho8gCEJWVpawd+9eYe3atcLq1auVDk2ysLAQbty4odFz/pOrV68Kw4YNE1q3bi20adNGGD58uHD16lWNxxEXFyfk5ORo/LxFuXz5stCsWTPBwMBA6ZBIJIKBgYFGY8nPzxdWrFgh9OzZU2jdurXQqlUrpUPTpFKp4ObmJkybNk1IS0vT+Plfmz17tmBubi6MHz9e+O9//yskJCQI48aNE8zNzYU5c+ZoLS4qe0yQSfD09BS2b98uCIIgWFpaionFuXPnBDs7O63EtGPHDqFr166CiYmJ4OHhIcTGxgp3797VSiyCIAgpKSnCiBEjhMqVKwsVKlQQhg4dKiQnJ2s8Dg8PD2HkyJE68Q96/fr1xUSv8N+b06dPC/b29hqPp3HjxsLs2bNV4jl58qTg6Oio8Xh06ZeZzp07C3FxcRo959vs2bNHMDExERo1aiSMGjVKCA8PFxo1aiRIpVJh3759Go3FwcFBsLKyEgYMGCAcP35co+d+U5MmTYTmzZsLv/zyi3D69GkhJSVF6dCkoUOHChYWFkJISIgwcuRIITw8XOnQtEePHgk//PCD4OvrKxgaGgpt27YVNm3aJMhkMo3G4eLiUuQvuPHx8YKLi4tGYyH1YoJMgqmpqXDz5k1BEJQTi8uXLwumpqbaDE14+PChMGfOHKFu3bqCkZGR8Omnnwpbt24V8vLytBKPXC4X5s2bJ0ilUsHAwECoW7eusHLlSkGhUGjk/Obm5uKfj7aZmZmJVcnCf2+uXbsmSKVSjcdjYWEhXL9+XSWeGzduaCUeXfplZsmSJYKDg4MwZswYYf369cJ///tfpUPT6tevL0yYMEFlfMKECYKvr69GY8nPzxf++9//Cl26dBFMTEyEWrVqCd9++62Qnp6u0TgE4dXPd2pqqsbPWxQ7Ozth9+7d2g6jSKdPnxaGDx8uVKpUSahYsaIwfPhwjf0CIZVKhStXrqiMX758WSv/nyH1YYJMgqenp5CQkCAIgnJi8cMPPwgNGjTQZmhK5s+fL0ilUkEikQiVK1cWJk+erLHkQy6XC5s2bRLatWsnGBoaCk2bNhVWrVolTJ8+XXBwcBB69uypkTi6dOkibNq0SSPn+idubm7C/v37BUFQ/nuzevVqwdPTU+PxVKtWTawAFo5n27Ztgpubm8bj0aVfZiQSSbGHpj+6F4RXScbly5dVxi9duqTVJOP+/fvC7NmzBR8fH8HY2Fjo2LGjkJCQIBQUFGjk/H5+flpvK3utatWqwqVLl7QdRrHu3r0rREZGClKpVLCwsBAMDQ2FZs2aCefPn1freevUqSPMmDFDZXzatGmCt7e3Ws9NmsUEmYRVq1YJ1apVEzZu3ChYWFgIGzZsEKZPny5+rU3p6enCzJkzhdq1awvm5uZC7969hUOHDgnr1q0TvL29hcDAQLWePzk5WRg2bJhgZ2cnVKlSRRgzZoxKhefkyZMaq7SvWLFCqFGjhhAZGSls2bJFq5XAmTNnCl5eXsKJEycEKysr4ejRo8K6deuEypUrCwsWLNBoLIIgCOPGjROaNWsmpKenC1ZWVsKVK1eEY8eOCW5ubkJUVJTG49GlX2Z0TfXq1YXNmzerjG/atElwcnLSQkT/c+LECeGLL74QpFKp4OLiItja2gouLi7Cb7/9pvZzHzx4UAgICBB+++034eHDh8LTp0+VDk36/vvvhSFDhmjs07GSkMvlws8//yy0b99eMDIyEho3biwsX75ceP78uZCWlib07NlT7b+cb9myRTA0NBSCgoKEqVOnCtOmTROCgoIEIyMjYdu2bWo9N2mWRBC4RwkBy5cvx/Tp08Ur3atVq4aoqCiEhYVpJZ5t27YhLi4Oe/fuhZeXFwYOHIg+ffrA1tZWnHPhwgX4+vpCLperLQ5DQ0MEBgYiLCwMwcHBMDY2VpmTk5ODYcOGIS4uTm1xvGZgUPzGMxKJBAUFBWqPobCJEydi7ty5ePnyJQBAKpVi7NixmDZtmkbjAIC8vDyEhoZi48aNEAQBRkZGKCgoQK9evRAfHw9DQ0O1x7Bjxw7x6wcPHmDq1Kno378/fHx8VP7udOrUSe3xFOXly5cwNTXVyrlfmzp1KubOnYuvv/4aTZo0gUQiwbFjxzBz5kyMGTMGkyZN0mg89+/fx9q1axEXF4fr168jODgYYWFhaNOmDXJzczFp0iRs2bJF3OFHXV7/fEskEqVxQRA08vPdtWtXpe8PHTqEihUrok6dOip/f7dt26bWWN40fPhwbNiwAQDQp08fDBw4EN7e3kpz0tLS4OLiAoVCodZYkpOTMXfuXKSmpkIQBHh5eWHMmDHw9fVV63lJs5ggk5KHDx9CoVCgSpUqWo3DxsYGPXv2RFhYGD788MMi5+Tm5mLWrFmIjIxUSwwFBQVYu3YtOnXqhIoVK6rlHO+DFy9e4OLFi1AoFPDy8oKlpaVW47l27RpOnz4NhUIBX19feHh4aOzcb/sFpjBN/zJTUFCAmJgYLFmyBPfv38fly5fh5uaGyZMnw8XFReO/CAuCgHnz5mH27Nm4d+8eAMDR0RHjxo3DiBEjVBJEderYsSP27t2LmjVrYuDAgejbt6/Kz/u9e/dQvXp1tSdeR44ceevjLVq0UOv5+/fvX+K5migIFNa6dWsMHDgQn332GUxMTIqck5+fj+PHj6v9fSL9wASZdE5+fj6WLVuGrl27wsHBQauxmJqaIjU1Fa6urlqNA3j1vpiamiIlJUWlcqJtt2/fhkQiQfXq1bUdChVh6tSpWL16NaZOnYpBgwbh/PnzcHNzw+bNmzF37lz88ccfWovt2bNnAAArKyutnD8sLAwDBw5EQEBAsXMEQUBaWhqcnZ01GBm9lpeXhy+++AKTJ0+Gm5ubVmMxNDREenq6ShHp0aNHqFKlisY/xSP1MdJ2AKQdvr6+Ja7S/PXXX2qORpmRkRHGjh2LTz/9VKPnLYqPjw+uX7+uEwmykZERnJ2ddeZ/wK9vQDF//nw8f/4cAGBpaYnhw4cjMjKyyHYUdRo9enSR4xKJBKampnB3d0fnzp318tOANWvWYNmyZWjdujW+/PJLcbxu3br4+++/NR7Pxx9/jG3btsHW1lYpMc7OzkZwcDAOHTqkkTjy8vJw/fp12NnZvXWeRCLRaHL84sULpKWlqbSP1a1bV2Mx5ObmQhAEmJubAwBu3bqF7du3w8vLC23bttVYHABgbGyM7du3Y/LkyRo9b1GKqynKZLJiK9tUPjFB1lPBwcHaDuGt/P39cfr0aa1XbGbMmCH21DZs2BAWFhZKj1tbW2s0nkmTJiEiIgLr1q3TeqI3bNgwbN++HbNmzRKrb3/88QeioqLw8OFDLFmyRKPxnD59Gn/99RcKCgpQq1YtCIKAK1euwNDQELVr18aiRYswZswYHDt2DF5eXmqPZ8SIEXB3d8eIESOUxhcuXIirV69i3rx5ao/htbt378Ld3V1lXKFQIC8vT2NxvHb48OEirx14+fIljh49qrE4jI2Ncf78eY22dLzNgwcP0L9/f/z6669FPq7JX447d+6Mrl274ssvv8STJ0/QqFEjmJiY4OHDh5gzZw6++uorjcUCAF26dEFCQkKxvwir2/z58wG8+mVpxYoVSq1kBQUF+P3331G7dm2txEbqwRYL0kk///wzvv76a4waNarIxFRTlZTCPaWF/xHV1EUzb/L19cXVq1eRl5cHZ2dnlfdFk9V+GxsbbNy4Ee3bt1ca//XXX9GjRw88ffpUY7EAr26NfvToUcTFxYm/uGRnZyMsLAzNmjXDoEGD0KtXL+Tm5mLv3r1qj6datWrYsWMHGjZsqDT+119/oVOnTrhz547aY3jNz88P4eHh6NOnD6ysrHDmzBm4ubkhOjoaBw4c0FhSevbsWQBA/fr1xQvAXisoKMCePXuwdOlS3Lx5UyPxAMCYMWNgbGyMb7/9VmPnLE7v3r1x8+ZNzJs3D61atcL27dtx//59TJ8+HbNnz9bop2qVKlXCkSNHUKdOHaxYsQILFizA6dOnsXXrVkyZMgWpqakaiwV4Vaz4/vvv0bp16yL/TXjzF9Gy9vpTxFu3bqF69epKF/2amJjAxcUFU6dOhb+/v1rjIM1hBZlEp06dQmpqKiQSCTw9PVX+Ydek7t27A1D+n55EItF4Yvrbb79p5DwlpUuVf1NTU7i4uKiMu7i4aOWjxu+++w779+9XqupbW1sjKioKbdu2xciRIzFlyhSNfTz86NEj2NjYqIxbW1vj4cOHGonhtcjISHz++ee4e/cuFAoFtm3bhkuXLmHNmjXYtWuXxuKoX78+JBIJJBIJPv74Y5XHzczMsGDBAo3FAwByuRwrVqzA/v374efnp5J4zZkzR2OxHDp0CP/973/x4YcfwsDAAM7OzggMDIS1tTViY2M1miC/ePFCbH/Zt28funbtCgMDAzRu3Fjtu3kUZcWKFbC1tUVycjKSk5OVHpNIJGpPkG/cuAEAaNWqFbZt24YKFSqo9XykfUyQCXfu3EHPnj1x/PhxcRu1J0+eoEmTJtiwYQOcnJw0HtPr/xlpU15eHqKiorB06VLUrFlT2+EAgNp27HgXQ4cOxbRp0xAXFwepVArgVR/ejBkzMGzYMI3H8/TpU2RmZqq0Tzx48ADZ2dkAAFtbW7VuC1iYu7s79uzZo/Je/Prrrxq/0Khjx47YtGkTYmJiIJFIMGXKFDRo0AA7d+5EYGCgxuK4ceMGBEGAm5sbTp48icqVK4uPmZiYoEqVKhrZjq+w8+fPo0GDBgCAy5cvKz2m6daLnJwc8eKvihUr4sGDB6hZsyZ8fHw0fi2Iu7s7EhIS0KVLF+zduxejRo0CAGRmZmq8tUwQBPz222+oUqWK2BOtLbpWNCH1YYJMGDBgAPLy8pCamopatWoBAC5duoQBAwYgLCwM+/bt02g8eXl5aNWqFXbt2qWRXtHi6Fp/YmHJyclitd/Ly0sr+2+ePn0aBw8eRPXq1VGvXj0AwJkzZyCXy9G6dWulPVU1sWdq586dMWDAAMyePRsffvghJBIJTp48ibFjx4qV95MnT2rsl53Ro0dj2LBhePDggVgtPXjwIGbPnq3R/uPXgoKCEBQUpPHzFvb6mgJ1b5dWUgUFBYiKioKPj4/We/oBoFatWrh06RJcXFxQv359LF26FC4uLliyZAmqVq2q0VimTJmCXr16YdSoUWjdurV4ncG+ffs0/v8bQRBQs2ZNXLhwQaPbNhbnzp072LFjR5EXUmryEwdSL/YgE8zMzP6vvTuPi7Jq/wf+mUEQkFXEHVmVzZUUKVNxYylFsDKVDPcsFFzTHhLI7aE019zCTNLMxN1MBfGBxA0VEBRQWRRcwBRxQ9nm+v3Bl/vHCJY9T3Nmkuv9es3rxdwzej4ONXPmvs+5Lpw8ebLOm15ycjJ69eqFp0+fCs/Upk0bHD16FI6OjsLHrk2T1icC1WdvRowYgfj4eJiYmICI8ODBA/Tr1w/bt29XOiOnappWM/Xx48eYPn06fvjhB1RWVgKorvwREBCA5cuXo0mTJkhNTQVQfalfhHXr1mHRokVSrV8rKyuEh4fjww8/FDL+8zRpGRUAZGRk1DvJENlERZNKOf74449Sw5uUlBR4enri3r170NHRwebNm6WlZ6IUFhbi9u3b6NKli7QfIykpCUZGRsI3pDk7O+O7776Dm5ub0HGfFxcXBx8fH1hbW+Py5cvo2LEjrl27BiKCi4uLsAosTPV4gsxgb2+PLVu2wNXVVel4UlISRo0ahezsbOGZIiIikJWVhY0bN6JRI/Vd6Jg6dSp++OEH2NnZqX19IlC9NjsnJwdbtmyRvjxkZGQgICAAdnZ2Uqephuzx48fIzc0FEcHW1lbtjUuA6mUeenp6asuiacuocnNz4efnh/T0dGlvAfD/lzSI3Pzao0cPREREYMCAAcLGfFmlpaXIyspCu3bt0KxZM3XHUauDBw8iIiIC69atU2sdeFdXV3h5eWH+/PnShtfmzZvD398fXl5ewqt7MNXhCTLDvn37sHjxYqxZswavvfYaZDIZzp07h6lTp2LOnDlq2Rjm5+eHuLg4GBgYoFOnTnUmpqLanPbr1++Fj8lkMuFnC4yNjXH06NE63QWTkpLg4eGBkpISoXnYP4OHhwcePnyIqKioOsuomjRpInwZ1ZAhQ6ClpYXIyEhpPfK9e/cwc+ZMLF26FL179xaWJSYmBnPmzNGYUo5A9cbBvLw82NraqvUEwdmzZxEdHV3vWX7RraZNTU1RWlqKyspK6OjoQE9PT+nx4uJiITkMDQ2RmpoKW1tbmJqaIjExEc7Ozrhw4QKGDh0qtAILUy1eg8wwZswYlJaWomfPntKbcWVlJRo1aoRx48Zh3Lhx0nNFvQmZmJjgnXfeETLWH9G0DRkKhaLeBhza2tpqWde5c+dO7Nixo94PUNGbip48eYKIiAjExcXhzp07dV6P3NxcoXkAzXl9jh8/jpMnT0qTY6D6ytHq1avRq1cvYTlqnDp1CseOHYO5uTnkcjnkcjnefPNN/Pvf/0ZQUBBSUlKEZfHy8gJQvaxD3aUcS0tLMXXqVERFRQGA1BI8KCgIrVu3xty5c4Vl2b59Oz788EN4eHggNjYWHh4euHr1KgoLC+Hn5ycsRw11rNuvT5MmTVBWVgaguj16Tk4OnJ2dAUB4dRqmWjxBZhrzxlObiDWrf0V2djZycnLQp08f6OnpSR+eovXv3x/BwcH46aef0Lp1awDVTSBqNtKItGrVKoSEhCAgIAD79u3D2LFjkZOTg7NnzyIwMFBoFgCYMGECEhISMHr0aLRq1Urtmys16fVp165dvQ1BKisr0aZNG6FZgOolFDXLTZo1a4Zbt27B3t4elpaWuHz5stAsmvQl+LPPPsOFCxcQHx8vTdwBYODAgQgLCxM6QV68eDGWL1+OwMBAGBoaYuXKlbC2tsZHH30kfMMgAAQEBAgfsz5ubm44ceIEnJyc8Pbbb2PmzJlIT0/H7t271b4+mv3NiDENVVFRQbGxsbR+/Xp6+PAhERHdvHmTHj16JCzD3bt3qX///iSTyUgul1NOTg4REY0bN45mzJghLEeN/Px86tatG2lra5ONjQ3Z2tqStrY2ubi4UEFBgdAs9vb2tG3bNiIiMjAwkF6befPmUWBgoNAsRETGxsaUmJgofNwX0aTXZ+/eveTq6kpnz54lhUJBRERnz54lNzc32rNnj9AsRERvvvmmNO7IkSPJy8uLEhMT6cMPPyRnZ2fheTRFu3bt6NSpU0Sk/N/M1atXydDQUGgWfX19ysvLIyIiMzMzSktLIyKijIwMatmypdAsNbKzsykkJIRGjBhBRUVFRER06NAhunjxorAMOTk5dOHCBSIievLkCX388cfUqVMn8vPzo2vXrgnLwVSPJ8iMiIgqKyspOjqa5s+fTwsWLKCdO3dSRUWF2vJcu3aNHBwcSF9fn7S0tKQPiuDgYProo4+E5Rg9ejR5enpSQUGB0gfWkSNHyMnJSViO58XExNCqVato5cqVFBsbq5YMenp60geCubk5paamEhHRlStXqGnTpsLzWFlZUUZGhvBxX0STXh8TExPS0dEhuVxOOjo6Sj+bmpoq3UQ4fPgw7dq1i4iqJxyOjo4kk8moWbNmdPToUSEZavvtt9/I39+fXn/9dbpx4wYREf3www90/PhxoTn09PSk95ja7zepqalkZGQkNEvbtm2lSXHnzp2lL3snT54UnoWIKD4+nvT09GjgwIGko6MjvTZffvklvfPOO8LzsFcfL7FguHjxIoYOHYrCwkJpjeKVK1dgbm6O/fv3o1OnTsIzBQcHo3v37rhw4QLMzMyk435+fpgwYYKwHDExMThy5Ajatm2rdLx9+/ZCu0kdO3YMU6ZMwenTp2FkZIRBgwZJDR4ePHgAZ2dnrF+/XujmppYtW+LevXuwtLSEpaUlTp8+jS5dukjNIERbsGABQkNDERUVpfZmAoBmvT6atoyqdj1mGxsbZGRkoLi4GKampsKXxuzatQujR4+Gv78/kpOTpfWljx49wuLFi/Hrr78Ky9KjRw8cPHgQU6dOBfD/q3pERkZKdYhF6d27N2JjY9GpUycMHz4cwcHBOHbsGGJjY9VS8WPu3LlYuHAhZsyYIXX4A6o3Uq9cuVJ4HqC6Ys7zex3UsamTqYi6Z+hM/Xr27ElDhgyh4uJi6VhxcTH5+PiQm5ubWjKZmZlRVlYWESmfScnLyyM9PT1hOQwMDOjKlSt1ciQlJQk9CzhkyBBatmzZCx9fuXIl+fr6CstDRDR+/HgKDw8nIqJ169ZJZ3dMTExo3LhxQrMQEXXt2pUMDQ3JwMCAOnbsSN26dVO6iaZpr48mGTt2rLRsqrbHjx/T2LFjhWbp2rUrRUVFEZHy/+MpKSnUokULoVlOnDhBhoaGNHnyZNLV1aXg4GAaOHAgNWnShM6dOyc0y7179+jmzZtERFRVVUVffvklDRkyhKZPn670WSFKkyZNKDc3l4jqfiY0btxYWI7c3Fx66623SF9fn+RyuXSrWYbHXh1c5o1BT08P586dk3bi1rh48SJ69OihlkYhTZs2RWJiIpycnKRakzY2NkhMTMQ777yDoqIiITnefvttuLi4YMGCBTA0NERaWhosLS0xYsQIKBQK7Ny5U0gOS0tLHD58+IWNU7KysuDh4YH8/HwheYDqihoKhUKqfLJjxw4kJibCzs4OkydPho6OjrAsAPDFF1/84eOi23Rr2utTVVWFPXv2KDUKGTp0qFrKiGlpaeH27dtSW+Uad+/eRcuWLaVGLyLo6+sjIyMDVlZWSu81ubm5cHJywrNnz4RlAYD09HQsXboU58+fh0KhgIuLC+bMmaOWK3mapG3bttixYwfeeOMNpd/Tnj17MGvWLOTk5AjJ8cYbbwCovsrZokWLOlc8+vbtKyQHUz1eYsFgb2+PoqKiOhPkO3fuwM7OTi2ZBg0ahBUrVuDbb78FUH2p8fHjxwgLC8Nbb70lLMeSJUvg7u6Oc+fOoby8HJ9++ikuXbqE4uJinDhxQliOoqKiesu71WjUqBF+//13YXkqKyuxaNEijBs3TmoyMXz4cAwfPlxYhueJngD/mZryZTXU+fpoyjKqhw8fgqr3vuDRo0fQ1dWVHquqqsKvv/5aZ9Ksaq1atUJ2djasrKyUjicmJsLGxkZoFgDo1KmTVOZNnf7sy3a7du0EJak2atQozJkzB9HR0ZDJZFAoFDhx4gRmzZoltDNlWloazp8/r1Qykb2i1HsCm2mCgwcPkrOzM0VHR1NBQQEVFBRQdHQ0derUiQ4ePEgPHjyQbqLcvHmTOnToQI6OjtSoUSNyc3MjMzMzsre3l3Yvi3L79m0KDQ2lt99+m7y9vSkkJIRu3bolNIONjQ3t3r37hY/v2rWLrK2tBSaqvuRZs8ud1a9m85ebm5taN39pyjKqmsvQL7ppaWnRwoULheUhqt7k5eTkRKdPnyZDQ0M6fvw4bd26lczNzWn16tVCsxBpzobpP/tdiVZeXk6jRo2SljNoa2uTXC6nDz74gCorK4XlcHd3V9vGaCYWL7FgSme5ai4X0XOtX0kNRfOfPn2Kn376CcnJydKlRn9//zodlBqCqVOnIj4+HmfPnlU66wZUv06urq7o168fVq1aJSyTr68vfH19MWbMGGFj/pGqqiosX778hY05RDW5qVF789eWLVuQkZEBGxsbrF27Fr/88ovQzV+asowqISEBRIT+/ftj165daNq0qfSYjo4OLC0tpfreIoWEhGD58uXScorGjRtj1qxZWLBggdAcmnKmHwAuXLigdL+iogIpKSlYtmwZFi1ahGHDhgnLUltOTg5SUlKgUCjQrVs3tG/fXvj4kydPxgcffICOHTvWubLXuXNnoXmY6vAEmSEhIeGFjyUnJ8PFxUW63xDXV5WUlCApKane7myiLu0VFRXBxcUFWlpamDJlCuzt7SGTyZCZmYk1a9agqqoKycnJaNGihZA8ALBhwwaEh4fD39+/3ha9Pj4+wrIAQGhoKDZu3IgZM2Zg3rx5CAkJwbVr17B3716EhoYiKChIaJ5u3bph+vTp+PDDD5XWTKampsLLywuFhYXCsnTt2hXLli1D//79lY4fO3YMwcHBSE9PF5YFAK5fvw4LCwulL+fqVlpaioyMDCgUCjg5OUmNTERyc3ND8+bNERUVBVNTUwDA/fv3MWbMGNy5cwenTp0Snul5Bw8exJIlSxAfH6+2DM+fwBHp9OnTGDVqlFJLaZlMppaTSEzF1HbummmskpISWrNmDXXr1k2tu3KzsrIoMDCQ+vfvTwMGDKDAwEDKzMwUmmH//v1kaGhIcrmcjI2NycTERLqJqhlb49q1a+Tt7S1dYqy5BOrt7a2WpQ41Geq7qeO/GxsbG/rll1+IqHqXe3Z2NhFVV/gYOXKk8Dx6enrS76X2rvucnByhu+6JNHMZ1f3792np0qU0fvx4mjBhAi1btoxKSkqEjV+f/Px84Q13atPV1a236UV6ejrp6uqqIVFdV65cIX19fbWMvXHjRnJ2dpZqeTs7O1NkZKTQDI6OjjRs2DA6ffo05eXl0bVr15Ru7NXBE2QmiYuLI39/f9LT0yMHBwcKCQmh5ORktWSJjo6W1h5Pnz6dpk+fTq+//jo1atSIduzYISxH+/btKTg4mJ48eSJszD9TXFxMSUlJdObMGbWUW9JU+vr6dP36dSIiatmyJZ0/f56Iqiek6mhsYGNjI61VrD1BjoqKIkdHR6FZnv/y8vyXLNFlqs6ePUtNmzalNm3akJ+fH/n6+lLbtm3JzMxM+r2JUlFRQZ9//jkZGRlJr4WRkRGFhIRQeXm50CxdunShuLi4Osfj4uKoY8eOQrPU/tL04MEDKikpoczMTHr//fepS5cuQrMQEX3++efUpEkTmjt3Lu3bt4/27dtHc+fOJQMDAwoJCRGWQ19fn65evSpsPKY+XMWigbtx4wY2b96MTZs24cmTJxg+fDgqKiqwa9cuODk5qS3Xp59+is8++wzz589XOh4WFoY5c+bgvffeE5Lj5s2bCAoK0ojGEzVMTU3Ro0cPtY3/9OlTxMXFYfDgwQCAzz77TGquAFRX1Jg/f36dtdKq1rZtW9y+fRvt2rWDnZ0dYmJi4OLigrNnz6Jx48ZCswDARx99hODgYGzatAkymQy3bt3CqVOnMGvWLISGhgrN8p///EfoeH9m+vTp8PHxQWRkpFRmrrKyEhMmTMC0adPw22+/CcsyZcoU7NmzB1999ZXUjOPUqVMIDw/H3bt3sX79epWO//DhQ+nnxYsXIygoCOHh4XBzcwNQfUl//vz5+PLLL1Wa43kmJiZ1ljAQESwsLLB9+3ahWQBg3bp1iIyMxMiRI6VjPj4+6Ny5M6ZOnYqFCxcKydG/f39cuHBBbRWemDi8BrkBe+utt5CYmIjBgwfD398fXl5e0NLSgra2Ni5cuKDWCbK+vj7S0tLqvAldvXoVXbp0QWlpqZAcw4YNw4gRI9RavkzTbNiwAb/88gsOHDgAADA0NISzs7O0eTIrKwuzZ8/GjBkzhOaaO3cujIyM8K9//Qs7d+7EyJEjYWVlhfz8fEyfPh0RERFC8wDA559/jmXLlql985em0dPTQ0pKChwcHJSOZ2RkoHv37sL+/wYAY2NjbN++Hd7e3krHDx06hBEjRuDBgwcqHV8ulytNRKmeDdI190Wub31+b4pcLoe5uTns7OzUUjvb1NQUSUlJdTblXblyBa6urigpKRGS49tvv8XChQsxbtw4dOrUqc4mPdF7L5jq8BnkBiwmJgZBQUH4+OOPhe8E/jPu7u44fvx4nQlyYmKiytsp79+/X/r57bffxuzZs5GRkcFvhv/nxx9/xPTp05WObdu2TaoZu3XrVqxZs0b4BLn2BPjdd9+FhYUFTpw4ATs7O6G/p9LSUsyePRt79+5FRUUFhgwZgpkzZwKA2jZ/AcDx48exYcMG5ObmIjo6Gm3atMGWLVtgbW2NN998U2gWIyMj5Ofn15kgFxQUKLURFkFXV7dODWQAsLKyEtLMRdPO7tfQtA3ZH3zwAdatW4dly5YpHf/222/h7+8vLMfkyZMBoM7VTUD8lximWjxBbsCOHz+OTZs2oXv37nBwcMDo0aPx/vvvqy1P7Ympj48P5syZg/PnzytdaoyOjv7Tjmn/K19f3zrH+M3w/7ty5Qo6dOgg3dfV1VWqRuDq6orAwEChmSoqKjBp0iTMmzdPmqj37NkTPXv2FJoDqF4GtHnzZqkk4bZt26BQKBAdHS08S43aJeeSk5OlJTGPHj3C4sWLhZacA4D3338f48ePx9KlS/HGG29AJpMhMTERs2fPVrqELkJgYCAWLFiA77//XlqKU1ZWhkWLFmHKlCkqH1/TJqI1ar8f/xlVfQGt/SVbJpNh48aNiImJUfpMKCgoENoo5PlKRuzVxUssGEpLS7F9+3Zs2rQJSUlJqKqqwrJlyzBu3DihZ3NetuRTQ52Yago9PT2kpqa+sJNUVlYWunbtKrxFr4mJCZKTk9XS/aw2W1tbLFq0CCNGjAAAJCUloVevXnj27Bm0tLTUkkmTSs4BkLpSrlu3Tmorra2tjY8//hgREREqXzP+fA3fo0ePonHjxujSpQuA6hrA5eXlGDBgAHbv3q3SLLVZW1tj7NixGDNmjPBOdc+rWfrx/BTh+WOqfD/u16/fSz1PJpPh2LFjKsnAGi7NKULJ1EZfXx/jxo1DYmIi0tPTMXPmTERERKB58+ZCL00rFIqXuomYHB87dgxOTk5KG2hqPHjwAM7Ozjh+/LjKc2iitm3b4uLFiy98PC0tDW3bthWYqJqfnx/27t0rfNznFRQUKC0DcnV1RaNGjXDr1i21Zbp8+TL69OlT57iRkZGwtZtA9ZfxwMBAWFtbY9u2bfD19UV8fDxSUlJQXFyM5cuXC9lQaWxsrHR75513MHjwYFhYWMDCwgKDBw/GsGHDYGxsrPIstc2YMQP79u2DjY0NBg0ahO3btyttgBUpJiYGXbt2xaFDh1BSUoIHDx7g0KFDcHFxwZEjR4S8H//nP/95qZvoyXFCQgKGDBkCOzs7tG/fHj4+Pg328+CVprb6GUyjVVZW0p49e2jIkCHqjkJPnz4VPuaQIUNo2bJlL3x85cqV5OvrKzCR5ggKCiInJ6d6fy+lpaXk5OREQUFBwnMtXLiQTExMaNiwYbR48WJauXKl0k0UuVxOd+7cUTpmYGBAubm5wjI8T1NKzs2aNYv09fVp4sSJFBQURM2aNaN3331X2Pj/FKmpqRQUFETm5uZkampKgYGBwsvfOTs719sS/bfffiMHBwehWYiIYmJiqLS0VPi4z9uyZQs1atSIhg8fTitXrqQVK1bQ8OHDSVtbm3788Ud1x2N/I15iwTRSVVUVFi9ejPXr16OoqAhXrlyBjY0N5s2bBysrK4wfP16l41taWuLw4cNwdHSs9/GsrCx4eHggPz9fpTk0UVFREbp27QodHR1MmTIFHTp0gEwmQ1ZWFr755htUVlYiJSVFaFc/oPry9IvIZDLk5uYKySGXy+Ht7a10JvTAgQPo37+/UrdBkZfuv/rqK0RFRWHTpk0YNGgQfv31V1y/fh3Tp09HaGiokLW2gGYuP6lx584dXL58GTKZDB06dEDz5s3VmgeoXlu/du1azJkzBxUVFejYsSOCg4MxduxYlXeR09PTQ1JSUp321mlpaejZs6ew9uQ1jIyMUFZWhtdeew19+/aFu7s7evXqJXzTq6OjIyZNmlRno/KyZcsQGRmJzMxMoXmY6vAEmWmk+fPnIyoqCvPnz8fEiRNx8eJF2NjYYMeOHVi+fLnKW67q6uri4sWLL6x1mZ2djU6dOgn/kNAUeXl5+PjjjxEbG6tUhmrQoEFYu3atWtcB3717FzKZDGZmZmoZf+zYsS/1vO+//17FSZRpQsk5HR0d5OXloU2bNtIxPT09XLlyBRYWFsJy1Pbw4UMEBgZi+/bt0nIBLS0tvP/++1izZo3wZRZA9cR4z549+P777xEbGws3NzeMHz8et27dwjfffIN+/fph27ZtKs3Qp08faGtrY+vWrWjVqhUAoLCwEKNHj0Z5eXmdMnCqVlVVhaSkJCQkJCA+Ph4nT57Es2fP4OLiAnd3d2FlHBs3boxLly7V+WzIzs5Gx44dhe+9YCqk1vPXjL2Ara0tHT16lIiULwlnZmaSiYmJyse3sbGh3bt3v/DxXbt2kbW1tcpzaLp79+7RmTNn6MyZM3Tv3j215bh//z598sknZGZmJnVDMzMzo8DAQLW3L1anJ0+e0CeffEKtW7cmc3NzGj58uPT7evTokfA8mrj85L333qP27dvT4cOH6cGDB/Tw4UM6fPgw2dvb03vvvSc0y/nz52nKlClkZmZGzZs3p5kzZ1JmZqbSc5KSkoS0nb569Sp17NiRtLW1ydbWlmxtbalRo0bk7OysEZ3k0tPTKSAggBo1aiS0tb2trS2tX7++zvH169eTnZ2dsBxM9bjMG9NIN2/erPfsrUKhQEVFhcrHf+uttxAaGgpvb+86HeGePn2KsLAwqZNcQ9a0aVO4urqqNUNxcTFef/113Lx5E/7+/nB0dAQRITMzE5s3b0ZcXBxOnjwJU1NTteZUB00rOUdEGDNmjNLyk2fPnmHy5MlqW35y8OBBHDlyRKkWtKenJyIjI+Hl5SUsBwD06NEDgwYNwrp16+Dr61un7jpQXUu7ZomKKtnZ2SEtLQ2xsbHIysoCEcHZ2RkDBgxQ+fKO+mRmZkpnjxMSElBVVYU333wTX3/9tZBSeePGjcPKlSsxc+ZMBAUFITU1ValE4ebNm7Fy5UqV52Di8BILppG6d++OadOm4YMPPlAqS/XFF1/g6NGjKt8xXFRUBBcXF2hpaWHKlCmwt7eHTCZDZmYm1qxZg6qqKiQnJwtfZ8vqmjZtGuLi4nD06NE6v4/CwkJ4eHhgwIABWL58uZoSqo+mrfnVxOUn7dq1w8GDB+tda/vWW2/hxo0bwrJcv34dlpaWwsarz5kzZ1BcXKzUWTAqKgphYWEoLS2Fr68vVq9eLbx9e00nv2nTpsHHxwfOzs5Cx9fS0sLt27fRvHlz7NmzB19//bW03tjR0RGzZ8/G0KFDhWZiKqbW89eMvcD+/fvJ2NiYIiIiSF9fn5YsWUITJkwgHR0diomJEZLh2rVr5O3tTXK5nGQyGclkMpLL5eTt7U15eXlCMrA/Z2lpSYcPH37h44cOHSJLS0txgTSItrY23bhxQ+mYrq4u5efnqymR5tmwYQMNHDiQbt26JR27ffs2eXh41HspXYRz587Rli1baOvWrcKrV3h5eVFERIR0Py0tjbS1tWnChAn09ddfU8uWLSksLExoJiKi4OBg6tatG+no6JCrqyt9+umn9OuvvwpbKiSTyaioqEjIWEwz8BlkprGOHDmCxYsX4/z581AoFHBxcUFoaCg8PDyE5rh//z6ys7NBRGjfvn2DvFSvyRo3boycnJwX1l6+ceMG7OzsGuTmGS0tLRQWFsLc3Fw6ZmhoiLS0tD+s+tGQdOvWDdnZ2SgrK5Oac+Tn56Nx48Zo37690nOTk5NVmuXOnTsYMWIE4uPjYWJiAiLCgwcP0K9fP2zfvl3p96gqrVq1woEDB9C9e3cAQEhICBISEpCYmAgAiI6ORlhYGDIyMlSepT4lJSU4fvw4EhISkJCQgPT0dHTt2hWnT59W6bhyuRxFRUVCfgdMM/AaZKaxPD094enpqe4YMDU1RY8ePdQdg71As2bNcO3atRdOkPPy8tRW0ULdSAPX/Gqa+lrLq8vUqVPx8OFDXLp0SSoxmZGRgYCAAAQFBeGnn35SeYb79+8rLVVKSEhQWovdo0cPFBQUqDzHiygUClRWVqK8vBxlZWWoqKjAtWvXhIxdU9LyjxQXFwvJwlSPzyAzxv7Rxo8fj+zsbMTGxkJHR0fpsbKyMnh6esLW1hbfffedmhKqjyau+WUvZmxsjKNHj9b5Qp6UlAQPDw8hXQ8tLS2xZcsW9OnTB+Xl5TAxMcGBAwcwYMAAAEB6ejr69u0rfCIYHByM+Ph4XLp0CU2bNkWfPn3g7u4Od3d3dOzYUeXjy+VyrFix4k/L/gUEBKg8CxODzyAzjSSXy//wm7qIdtPsn+GLL75A9+7d0b59ewQGBsLBwQFA9Zm3tWvXoqysDFu2bFFzSvXgie+fKygogEwmk65AJCUlYdu2bXBycsKkSZOEZlEoFPVWrtDW1oZCoRCSwcvLC3PnzsWXX36JvXv3Ql9fX6l1elpaGmxtbYVkqe3mzZuYOHGisAlxfUaMGKERDWSYGHwGmWmkffv2Kd2vqKhASkoKoqKi8MUXX6i8kx77Z8nLy8Mnn3yCmJiYOo1Lvvnmmxc2fGGsd+/emDRpEkaPHo3CwkJ06NABHTt2xJUrVxAUFITQ0FBhWYYOHYqSkhL89NNPaN26NQBI5QtNTU2xZ88elWf4/fffMWzYMJw4cQIGBgaIioqCn5+f9PiAAQPg5uaGRYsWqTyLJqldxYI1DDxBZv8o27Ztw88//1xnAs0YUL1+8urVqwCq67g2bdpUzYmYpjM1NcXp06dhb2+PVatW4eeff8aJEycQExODyZMnC2tRDlSfzR46dCguXrwICwsLyGQyXL9+HZ07d8a+ffteuM5eFR48eAADA4M65QCLi4thYGBQZzmTKBkZGcjPz0d5ebnScR8fH5WOK5fLUVhYyBPkBoQnyOwfJScnB507d8aTJ0/UHYUx9gowMDDAxYsXYWVlBR8fH/Tq1Qtz5sxBfn4+7O3t1dJOvr7mHA1dbm4u/Pz8kJ6eDplMpnSlCOBld+zvJ1d3AMZe1tOnT7F69WqhZ1EYY682Z2dnrF+/HsePH0dsbKxUseHWrVvCqp+cOXMGhw4dku4PGjQIRkZGWLZsGUaOHIlJkyahrKxMSBZNFRwcDGtraxQVFUFfXx+XLl3Cb7/9hu7duyM+Pl7d8dgriDfpMY1kamqqtEmPiPDo0SPo6+tj69atakzGGHuVfPnll/Dz88OSJUsQEBCALl26AAD2798vrI16eHg43N3dpe516enpmDhxIgICAuDo6IglS5agdevWCA8PF5JHE506dQrHjh2Dubk55HI55HI53nzzTfz73/9GUFAQUlJS1B2RvWJ4gsw00vLly5UmyDVtRnv27MmNOhhjfxt3d3fcvXsXDx8+VHpvmTRpEvT19YVkSE1NxYIFC6T727dvh6urKyIjIwEAFhYWCAsLa9AT5KqqKhgYGACorn1+69Yt2Nvbw9LSEpcvX1ZzOvYq4gky00hjxoxRdwTGWAOhpaVV54u3lZWVsPE1vTmHJujYsSPS0tJgY2ODnj174quvvoKOjg6+/fZb2NjYqDseewXxBJlpjLS0tJd+bufOnVWYhDHWkOzcuRM7duyotzqCqttLA0CLFi2Ql5cHCwsLlJeXIzk5GV988YX0+KNHj+qtj9yQfP7559Lm7IULF2Lw4MHo3bs3zMzM8PPPP6s5HXsV8QSZaYyuXbsq7U5+EZlMxjuWGWN/i1WrViEkJAQBAQHYt28fxo4di5ycHJw9exaBgYFCMmhqcw5N4unpKf1sY2ODjIwMFBcX19mvwtjfhcu8MY1x/fr1l36upaWlCpMwxhoKBwcHhIWFYeTIkTA0NMSFCxdgY2OD0NBQFBcX45tvvlF5Bm7O8ccqKyuhq6uL1NRUtXXRYw0PT5CZRrp3755UYqmgoACRkZF4+vQpfHx8lM6sMMbY/0JfXx+ZmZmwtLRE8+bNERsbiy5duuDq1atwc3PDvXv3hGXR1OYcmsDW1ha7d++WqowwpmpcB5lplPT0dFhZWaF58+ZwcHBAamoqevTogeXLl+Pbb79Fv379sHfvXnXHZIy9Ilq2bClNgi0tLXH69GkA1e3LRZ8/MjY2rjM5BoCmTZs26MkxUL0G+bPPPkNxcbG6o7AGgs8gM43i7e2NRo0aYc6cOdi6dSt++eUXeHh4YOPGjQCAqVOn4vz589KHGGOM/S8mTJgglVFbv349ZsyYgV69euHcuXMYNmwYvvvuO3VHZAC6deuG7OxsVFRUwNLSEk2aNFF6XMRmStaw8ASZaZRmzZrh2LFj6Ny5Mx4/fgwjIyMkJSWhe/fuAICsrCy4ubmhpKREvUEZY68EhUIBhUKBRo2q96zv2LEDiYmJsLOzg5+fHywsLNSckAHVzVT+aDNeWFiYwDSsIeAJMtMocrkchYWFaN68OQAobZoBgKKiIrRu3ZqrWDDGVKawsBCLFi3Cxo0b8fTpU3XHYYypAa9BZhrn+bMEXMKHMfZ3Kykpgb+/P8zNzdG6dWusWrUKCoUCoaGhsLW1xenTp7Fp0yZ1x2T/x8bGpt4NkyUlJdwohKkE10FmGmfMmDFo3LgxAODZs2eYPHmytN6srKxMndEYY6+If/3rX/jtt98QEBCAw4cPY/r06Th8+DCePXuGX3/9FX379lV3RFbLtWvX6r1yWFZWhhs3bqghEXvV8QSZaZSAgACl+x988EGd53z44Yei4jDGXlEHDx7E999/j4EDB+KTTz6BnZ0dOnTogBUrVqg7Gqtl//790s9HjhyBsbGxdL+qqgpxcXGwtrZWRzT2iuM1yIwxxhocbW1tXL9+Ha1btwZQXQ85KSmJG1FoGLm8eiVofV1WtbW1YWVlha+//hqDBw9WRzz2CuMzyIwxxhochUIBbW1t6b6Wllad0mFM/RQKBQDA2toaZ8+eRbNmzdSciDUUPEFmjDHW4BDRH+53qLF79251xGP/58yZMyguLkZeXp507IcffkBYWBiePHkCX19frF69Wvo9MvZ34QkyY4yxBudl9jsw9QsLC0O/fv3g7e0NoLrb6vjx4zFmzBg4OjpiyZIlaN26NcLDw9UblL1yeA0yY4wxxjRSq1atcODAAalZVEhICBISEpCYmAgAiI6ORlhYGDIyMtQZk72CuA4yY4wxxjTS/fv30aJFC+l+QkICvLy8pPs9evRAQUGBOqKxVxxPkBljjDGmkVq0aCGtPy4vL0dycjJef/116fFHjx4pbbZk7O/CE2TGGGOMaSQvLy/MnTsXx48fx2effQZ9fX307t1bejwtLQ22trZqTMheVbxJjzHGGGMaaeHChRg2bBj69u0LAwMDREVFQUdHR3p806ZN8PDwUGNC9qriTXqMMcYY02gPHjyAgYEBtLS0lI4XFxfDwMBAadLM2N+BJ8iMMcYYY4zVwmuQGWOMMcYYq4UnyIwxxhhjjNXCE2TGGGOMMcZq4QkyY4xpOHd3d0ybNu2lnx8fHw+ZTIaSkpL/aVwrKyusWLHif/o7GGPsn4gnyIwxVg+ZTPaHtzFjxqg7ImOMMRXhOsiMMVaP27dvSz///PPPCA0NxeXLl6Vjenp6Ss+vqKjgjl6MMfaK4DPIjDFWj5YtW0o3Y2NjyGQy6f6zZ89gYmKCHTt2wN3dHbq6uti6dSvCw8PRtWtXpb9nxYoVsLKyUjr2/fffw9HREbq6unBwcMDatWv/UratW7eie/fuMDQ0RMuWLTFq1CjcuXOnzvNOnDiBLl26QFdXFz179kR6errS4ydPnkSfPn2gp6cHCwsLBAUF4cmTJy8cNzw8HO3atUPjxo3RunVrBAUF/aXcjDH2T8ETZMYY+y/NmTMHQUFByMzMhKen50v9mcjISISEhGDRokXIzMzE4sWLMW/ePERFRb30uOXl5ViwYAEuXLiAvXv3Ii8vr94lH7Nnz8bSpUtx9uxZNG/eHD4+PqioqAAApKenw9PTE8OGDUNaWhp+/vlnJCYmYsqUKfWOuXPnTixfvhwbNmzA1atXsXfvXnTq1OmlMzPG2D8JL7FgjLH/0rRp0zBs2LC/9GcWLFiAr7/+Wvpz1tbWyMjIwIYNGxAQEPBSf8e4ceOkn21sbLBq1Sq4urri8ePHMDAwkB4LCwvDoEGDAABRUVFo27Yt9uzZg+HDh2PJkiUYNWqUtPmvffv2WLVqFfr27Yt169ZBV1dXacz8/Hy0bNkSAwcOhLa2Ntq1awdXV9e/9G9njLF/Cj6DzBhj/6Xu3bv/pef//vvvKCgowPjx42FgYCDdFi5ciJycnJf+e1JSUjB06FBYWlrC0NAQ7u7uAKonsbW9/vrr0s9NmzaFvb09MjMzAQDnz5/H5s2blXJ4enpCoVAgLy+vzpjvvfcenj59ChsbG0ycOBF79uxBZWXlX/r3M8bYPwWfQWaMsf9SkyZNlO7L5XIQkdKxmiUNAKBQKABUL7Po2bOn0vO0tLReaswnT57Aw8MDHh4e2Lp1K8zNzZGfnw9PT0+Ul5f/6Z+XyWRSlo8++qjedcTt2rWrc8zCwgKXL19GbGwsjh49ik8++QRLlixBQkICb05kjL1yeILMGGN/E3NzcxQWFoKIpIloamqq9HiLFi3Qpk0b5Obmwt/f/78aIysrC3fv3kVERAQsLCwAAOfOnav3uadPn5Ymu/fv38eVK1fg4OAAAHBxccGlS5dgZ2f30mPr6enBx8cHPj4+CAwMhIODA9LT0+Hi4vJf/VsYY0xT8QSZMcb+Ju7u7vj999/x1Vdf4d1338Xhw4dx6NAhGBkZSc8JDw9HUFAQjIyM4O3tjbKyMpw7dw7379/HjBkz/nSMdu3aQUdHB6tXr8bkyZNx8eJFLFiwoN7nzp8/H2ZmZmjRogVCQkLQrFkz+Pr6AqjeYOjm5obAwEBMnDgRTZo0QWZmJmJjY7F69eo6f9fmzZtRVVWFnj17Ql9fH1u2bIGenh4sLS3/uxeLMcY0GK9BZoyxv4mjoyPWrl2LNWvWoEuXLkhKSsKsWbOUnjNhwgRs3LgRmzdvRqdOndC3b19s3rwZ1tbWLzWGubk5Nm/ejOjoaDg5OSEiIgJLly6t97kREREIDg7Ga6+9htu3b2P//v3Q0dEBAHTu3BkJCQm4evUqevfujW7dumHevHlo1apVvX+XiYkJIiMj0atXL3Tu3BlxcXE4cOAAzMzM/sIrxBhj/wwyen7BHGOMMcYYYw0Yn0FmjDHGGGOsFp4gM8YYY4wxVgtPkBljjDHGGKuFJ8iMMcYYY4zVwhNkxhhjjDHGauEJMmOMMcYYY7XwBJkxxhhjjLFaeILMGGOMMcZYLTxBZowxxhhjrBaeIDPGGGOMMVYLT5AZY4wxxhirhSfIjDHGGGOM1fL/APAGptHla0jfAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<Figure size 800x600 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Compute the confusion matrix\n",
    "cm = confusion_matrix(valid_gen.classes, np.argmax(predictions, axis=-1))\n",
    "\n",
    "#Compute the classification metrics\n",
    "accuracy = accuracy_score(valid_gen.classes, np.argmax(predictions, axis=-1))\n",
    "precision = precision_score(valid_gen.classes, np.argmax(predictions, axis=-1), average='macro')\n",
    "recall = recall_score(valid_gen.classes, np.argmax(predictions, axis=-1), average='macro')\n",
    "f1 = f1_score(valid_gen.classes, np.argmax(predictions, axis=-1), average='macro')\n",
    "print('Accuracy:',accuracy.round(4))\n",
    "print('Precision:',precision.round(4))\n",
    "print('Recall:',recall.round(4))\n",
    "print('F1:',f1.round(4))\n",
    "\n",
    "# Plot the confusion matrix\n",
    "plt.figure(figsize=(8,6))\n",
    "sns.heatmap(cm.T, xticklabels=list(valid_gen.class_indices), yticklabels=list(valid_gen.class_indices))\n",
    "plt.xlabel('True labels')\n",
    "plt.ylabel('Predicted labels')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "ccbfea68-581a-4852-8e6d-41cb9f0d34dd",
   "metadata": {
    "collapsed": true,
    "jupyter": {
     "outputs_hidden": true
    },
    "tags": []
   },
   "outputs": [
    {
     "ename": "NameError",
     "evalue": "name 'X' is not defined",
     "output_type": "error",
     "traceback": [
      "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
      "\u001b[0;31mNameError\u001b[0m                                 Traceback (most recent call last)",
      "\u001b[0;32m<ipython-input-47-b4180b7044ae>\u001b[0m in \u001b[0;36m<module>\u001b[0;34m\u001b[0m\n\u001b[1;32m     12\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     13\u001b[0m     \u001b[0mnewImage\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0maug_test_data_gen\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mapply_transform\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m---> 14\u001b[0;31m     \u001b[0mpreds\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mmodel\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mpredict\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mX\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m     15\u001b[0m     \u001b[0mprint\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m'Before '\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mpreds\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mshape\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     16\u001b[0m     \u001b[0mpreds_tta\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mappend\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mpreds\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;31mNameError\u001b[0m: name 'X' is not defined"
     ]
    }
   ],
   "source": [
    "from tensorflow.keras.preprocessing.image import ImageDataGenerator\n",
    "\n",
    "tta_steps = 5\n",
    "preds_tta = []\n",
    "for i in range(tta_steps):\n",
    "    aug_test_data_gen = ImageDataGenerator(rotation_range=30,\n",
    "                                           height_shift_range=50,\n",
    "                                           width_shift_range=50,\n",
    "                                           zoom_range=0.3,\n",
    "                                           horizontal_flip=True,\n",
    "                                           vertical_flip=True, \n",
    "                                           fill_mode='reflect',\n",
    "                                           rescale=1/255.)\n",
    "    \n",
    "    gen = ImageDataGenerator(fill_mode='constant', cval=0.)\n",
    "\n",
    "    newX = gen.apply_transform(X, aug_test_data_gen)\n",
    "    preds = model.predict(newX)\n",
    "    #print('Before ', preds.shape)\n",
    "    preds_tta.append(preds)\n",
    "    #print(i,  len(preds_tta))\n",
    "final_pred = np.mean(preds_tta, axis=0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "9933ad4c-b524-422d-b981-17d7158d4e7e",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 49,
   "id": "5228fddb-56ca-409f-9d20-3bf649601f4d",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 5/5 [01:02<00:00, 12.56s/it]\n"
     ]
    }
   ],
   "source": [
    "tta_steps = 5\n",
    "preds_tta = []\n",
    "for i in tqdm(range(tta_steps)):\n",
    "    preds = model.predict(valid_gen)\n",
    "    #     print('Before ', preds.shape)\n",
    "    preds_tta.append(preds)\n",
    "    #     print(i,  len(preds_tta))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "id": "feba18ca-feb2-4ed2-8e24-55a3d5434e2f",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(1779, 14)"
      ]
     },
     "execution_count": 25,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "final_pred = np.mean(preds_tta, axis=0)\n",
    "predicted_class_indices = np.argmax(final_pred, axis=1)\n",
    "final_pred.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "id": "5f038df0-c954-404e-8218-b07539b5650b",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(1779, 14)"
      ]
     },
     "execution_count": 27,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Predict the test set with the CNN\n",
    "predictions = model.predict(valid_gen)\n",
    "predictions.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 53,
   "id": "74745f3c-1d61-473e-900b-f141f1c9653b",
   "metadata": {},
   "outputs": [
    {
     "ename": "NameError",
     "evalue": "name 'final_pred' is not defined",
     "output_type": "error",
     "traceback": [
      "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
      "\u001b[0;31mNameError\u001b[0m                                 Traceback (most recent call last)",
      "\u001b[0;32m<ipython-input-53-2ce6d8d7832e>\u001b[0m in \u001b[0;36m<module>\u001b[0;34m\u001b[0m\n\u001b[1;32m      1\u001b[0m \u001b[0;31m# Compute the confusion matrix\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m----> 2\u001b[0;31m \u001b[0mcm\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mconfusion_matrix\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mvalid_gen\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mclasses\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mnp\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0margmax\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mfinal_pred\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0maxis\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;34m-\u001b[0m\u001b[0;36m1\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m      3\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m      4\u001b[0m \u001b[0;31m#Compute the classification metrics\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m      5\u001b[0m \u001b[0maccuracy\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0maccuracy_score\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mvalid_gen\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mclasses\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mnp\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0margmax\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mfinal_pred\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0maxis\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;34m-\u001b[0m\u001b[0;36m1\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;31mNameError\u001b[0m: name 'final_pred' is not defined"
     ]
    }
   ],
   "source": [
    "# Compute the confusion matrix\n",
    "cm = confusion_matrix(valid_gen.classes, np.argmax(final_pred, axis=-1))\n",
    "\n",
    "#Compute the classification metrics\n",
    "accuracy = accuracy_score(valid_gen.classes, np.argmax(final_pred, axis=-1))\n",
    "precision = precision_score(valid_gen.classes, np.argmax(final_pred, axis=-1), average='macro')\n",
    "recall = recall_score(valid_gen.classes, np.argmax(final_pred, axis=-1), average='macro')\n",
    "f1 = f1_score(valid_gen.classes, np.argmax(final_pred, axis=-1), average='macro')\n",
    "print('Accuracy:',accuracy.round(4))\n",
    "print('Precision:',precision.round(4))\n",
    "print('Recall:',recall.round(4))\n",
    "print('F1:',f1.round(4))\n",
    "\n",
    "# Plot the confusion matrix\n",
    "plt.figure(figsize=(8,6))\n",
    "sns.heatmap(cm.T, xticklabels=list(valid_gen.class_indices), yticklabels=list(valid_gen.class_indices))\n",
    "plt.xlabel('True labels')\n",
    "plt.ylabel('Predicted labels')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "id": "c060c789-9010-4d55-ac0a-cef0920393d5",
   "metadata": {
    "jupyter": {
     "source_hidden": true
    },
    "tags": []
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Accuracy: 0.9567\n",
      "Precision: 0.9522\n",
      "Recall: 0.9469\n",
      "F1: 0.9489\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 800x600 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Compute the confusion matrix\n",
    "cm = confusion_matrix(valid_gen.classes, np.argmax(predictions, axis=-1))\n",
    "\n",
    "#Compute the classification metrics\n",
    "accuracy = accuracy_score(valid_gen.classes, np.argmax(predictions, axis=-1))\n",
    "precision = precision_score(valid_gen.classes, np.argmax(predictions, axis=-1), average='macro')\n",
    "recall = recall_score(valid_gen.classes, np.argmax(predictions, axis=-1), average='macro')\n",
    "f1 = f1_score(valid_gen.classes, np.argmax(predictions, axis=-1), average='macro')\n",
    "print('Accuracy:',accuracy.round(4))\n",
    "print('Precision:',precision.round(4))\n",
    "print('Recall:',recall.round(4))\n",
    "print('F1:',f1.round(4))\n",
    "\n",
    "# Plot the confusion matrix\n",
    "plt.figure(figsize=(8,6))\n",
    "sns.heatmap(cm.T, xticklabels=list(valid_gen.class_indices), yticklabels=list(valid_gen.class_indices))\n",
    "plt.xlabel('True labels')\n",
    "plt.ylabel('Predicted labels')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "id": "6326332b-65b3-4496-99fb-cada8670ecf6",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "/Users/irfanmasudi/ANNDL/Challenge/datasets/train\n"
     ]
    }
   ],
   "source": [
    "%cd /Users/irfanmasudi/ANNDL/Challenge/datasets/train"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "id": "8b341281-be62-4515-a790-dc30127a2dc1",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>class</th>\n",
       "      <th>count</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>Apple</td>\n",
       "      <td>889</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>Blueberry</td>\n",
       "      <td>420</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>Cherry</td>\n",
       "      <td>524</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>Corn</td>\n",
       "      <td>1085</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>Grape</td>\n",
       "      <td>1312</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>Orange</td>\n",
       "      <td>1573</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>Peach</td>\n",
       "      <td>879</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>Pepper</td>\n",
       "      <td>688</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>Potato</td>\n",
       "      <td>644</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>Raspberry</td>\n",
       "      <td>237</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>Soybean</td>\n",
       "      <td>1454</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>Squash</td>\n",
       "      <td>516</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12</th>\n",
       "      <td>Strawberry</td>\n",
       "      <td>605</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13</th>\n",
       "      <td>Tomato</td>\n",
       "      <td>870</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "         class  count\n",
       "0        Apple    889\n",
       "1    Blueberry    420\n",
       "2       Cherry    524\n",
       "3         Corn   1085\n",
       "4        Grape   1312\n",
       "5       Orange   1573\n",
       "6        Peach    879\n",
       "7       Pepper    688\n",
       "8       Potato    644\n",
       "9    Raspberry    237\n",
       "10     Soybean   1454\n",
       "11      Squash    516\n",
       "12  Strawberry    605\n",
       "13      Tomato    870"
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "sample = []\n",
    "for label in labels:\n",
    "    count = 0\n",
    "    dir = label\n",
    "    for path in os.listdir(dir):\n",
    "        if os.path.isfile(os.path.join(dir, path)):\n",
    "            count += 1\n",
    "    sample.append([label,count])\n",
    "sample = pd.DataFrame(sample)\n",
    "sample.columns=['class','count']\n",
    "sample"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "id": "429fe06c-4714-4df3-af2f-a9796ae5fefe",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "sample.plot(x='class', kind='bar')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "id": "8973bcc0-5d9e-4c14-8997-92b8cf23c9c7",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Confusion Matrix\n",
      "[[ 83   1   1   0   2   1   0   3   0   0   2   1   1   4]\n",
      " [  0  46   1   0   0   0   0   0   0   0   0   0   0   0]\n",
      " [  0   0  56   0   0   0   0   1   0   0   1   0   0   1]\n",
      " [  3   0   0 116   0   0   0   0   0   0   0   1   0   1]\n",
      " [  0   0   2   0 138   0   0   0   1   0   0   1   1   3]\n",
      " [  0   1   2   0   0 168   0   0   0   0   1   0   0   3]\n",
      " [  1   1   0   0   0   0  89   4   0   0   0   0   0   3]\n",
      " [  1   0   0   0   0   1   0  73   0   0   1   0   0   1]\n",
      " [  0   0   0   0   0   0   0   0  70   0   1   0   0   1]\n",
      " [  0   0   0   0   0   0   0   0   0  26   0   0   1   0]\n",
      " [  1   0   0   0   0   0   0   0   0   0 160   0   0   1]\n",
      " [  1   0   0   2   0   0   0   0   0   0   0  52   0   3]\n",
      " [  0   0   0   0   0   0   0   0   0   0   0   0  66   2]\n",
      " [  0   1   0   0   0   5   1   3   1   0   0   0   0 559]]\n"
     ]
    }
   ],
   "source": [
    "y_pred = np.argmax(predictions, axis=1)\n",
    "print('Confusion Matrix')\n",
    "print(confusion_matrix(valid_gen.classes, y_pred))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "fbf9465e-bc34-4398-aef3-b7c4f7b8f35a",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Confusion Matrix\n",
      "[[ 90   0   2   0   1   0   0   3   0   0   0   1   0   2]\n",
      " [  0  47   0   0   0   0   0   0   0   0   0   0   0   0]\n",
      " [  0   0  57   0   0   0   0   0   0   0   1   0   0   1]\n",
      " [  1   1   0 119   0   0   0   0   0   0   0   0   0   0]\n",
      " [  0   0   0   0 141   0   0   0   1   0   0   1   2   1]\n",
      " [  0   1   0   0   0 173   0   1   0   0   0   0   0   0]\n",
      " [  0   0   0   0   0   0  94   1   0   0   0   0   0   3]\n",
      " [  0   0   0   0   0   0   0  73   0   0   2   0   0   2]\n",
      " [  0   0   0   0   0   0   0   0  68   0   3   0   0   1]\n",
      " [  0   0   0   0   0   0   0   0   0  27   0   0   0   0]\n",
      " [  1   0   0   0   0   0   0   0   0   0 160   0   0   1]\n",
      " [  0   0   0   2   1   0   0   0   0   0   0  52   0   3]\n",
      " [  0   0   0   0   1   0   0   0   0   0   0   0  66   1]\n",
      " [  0   0   0   0   0   0   0   0   1   0   0   0   0 569]]\n"
     ]
    }
   ],
   "source": [
    "y_pred = np.argmax(final_pred, axis=1)\n",
    "print('Confusion Matrix')\n",
    "print(confusion_matrix(valid_gen.classes, y_pred))"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3.9 (tensorflow)",
   "language": "python",
   "name": "tensorflow"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.9.5"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}