Improved SE on FER-2013.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "id": "04354bd1",
   "metadata": {},
   "source": [
    "### Improved SE on FER-2013\n",
    "\n",
    "See the other notebooks for the simpler models and for the other dataset."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "3bc308fc",
   "metadata": {},
   "outputs": [],
   "source": [
    "from tensorflow.keras.models import Model\n",
    "from tensorflow.keras.layers import Input, Conv2D, MaxPooling2D, Dense, Dropout, Flatten, BatchNormalization, GlobalAveragePooling2D, ReLU, Multiply, Reshape\n",
    "from tensorflow.keras.optimizers import Adam\n",
    "from tensorflow.keras.optimizers.schedules import ExponentialDecay\n",
    "from tensorflow.keras.preprocessing.image import ImageDataGenerator\n",
    "from tensorflow.keras.callbacks import ModelCheckpoint, EarlyStopping, TensorBoard\n",
    "\n",
    "import matplotlib.pyplot as plt\n",
    "import numpy as np"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "cec400a4",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Found 22968 images belonging to 7 classes.\n",
      "Found 5741 images belonging to 7 classes.\n",
      "Found 7178 images belonging to 7 classes.\n"
     ]
    }
   ],
   "source": [
    "# Enhanced train_data_gen with data augmentation; adapted from https://github.com/hujie-frank/SENet \n",
    "\n",
    "train_data_gen = ImageDataGenerator(\n",
    "    rescale=1./255,\n",
    "    validation_split=0.2,  # For validation data\n",
    "    horizontal_flip=True,  # Random mirroring\n",
    "    rotation_range=10,  # Random rotation between -10 to 10 degrees\n",
    "    width_shift_range=0.08,  # Random horizontal shift, 8% of total width\n",
    "    height_shift_range=0.08,  # Random vertical shift, 8% of total height\n",
    "    brightness_range=[0.8, 1.2],  # Pixel jitter for brightness\n",
    "    zoom_range=[0.92, 1.08],  # Random zoom, simulating cropping\n",
    "    fill_mode='nearest'  # How to fill points outside the boundaries\n",
    ")\n",
    "\n",
    "train_generator = train_data_gen.flow_from_directory(\n",
    "    'datasets/FER-2013/train',\n",
    "    target_size=(48, 48),\n",
    "    batch_size=32,\n",
    "    color_mode=\"grayscale\",\n",
    "    class_mode='categorical',\n",
    "    subset='training')\n",
    "\n",
    "validation_generator = train_data_gen.flow_from_directory(\n",
    "    'datasets/FER-2013/train', \n",
    "    target_size=(48, 48),\n",
    "    batch_size=32,\n",
    "    color_mode=\"grayscale\",\n",
    "    class_mode='categorical',\n",
    "    subset='validation')\n",
    "\n",
    "test_data_gen = ImageDataGenerator(rescale=1./255) # normalize\n",
    "\n",
    "test_generator = test_data_gen.flow_from_directory(\n",
    "    'datasets/FER-2013/test',\n",
    "    target_size=(48, 48),\n",
    "    batch_size=32,\n",
    "    color_mode=\"grayscale\",\n",
    "    class_mode='categorical')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "8cf43562",
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "2024-03-22 01:54:58.589309: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1639] Created device /job:localhost/replica:0/task:0/device:GPU:0 with 10398 MB memory:  -> device: 0, name: NVIDIA GeForce GTX 1080 Ti, pci bus id: 0000:04:00.0, compute capability: 6.1\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model: \"model\"\n",
      "__________________________________________________________________________________________________\n",
      " Layer (type)                Output Shape                 Param #   Connected to                  \n",
      "==================================================================================================\n",
      " input_1 (InputLayer)        [(None, 48, 48, 1)]          0         []                            \n",
      "                                                                                                  \n",
      " conv2d (Conv2D)             (None, 48, 48, 32)           320       ['input_1[0][0]']             \n",
      "                                                                                                  \n",
      " re_lu (ReLU)                (None, 48, 48, 32)           0         ['conv2d[0][0]']              \n",
      "                                                                                                  \n",
      " batch_normalization (Batch  (None, 48, 48, 32)           128       ['re_lu[0][0]']               \n",
      " Normalization)                                                                                   \n",
      "                                                                                                  \n",
      " global_average_pooling2d (  (None, 32)                   0         ['batch_normalization[0][0]'] \n",
      " GlobalAveragePooling2D)                                                                          \n",
      "                                                                                                  \n",
      " dense (Dense)               (None, 2)                    64        ['global_average_pooling2d[0][\n",
      "                                                                    0]']                          \n",
      "                                                                                                  \n",
      " dense_1 (Dense)             (None, 32)                   64        ['dense[0][0]']               \n",
      "                                                                                                  \n",
      " reshape (Reshape)           (None, 1, 1, 32)             0         ['dense_1[0][0]']             \n",
      "                                                                                                  \n",
      " multiply (Multiply)         (None, 48, 48, 32)           0         ['batch_normalization[0][0]', \n",
      "                                                                     'reshape[0][0]']             \n",
      "                                                                                                  \n",
      " conv2d_1 (Conv2D)           (None, 48, 48, 64)           18496     ['multiply[0][0]']            \n",
      "                                                                                                  \n",
      " re_lu_1 (ReLU)              (None, 48, 48, 64)           0         ['conv2d_1[0][0]']            \n",
      "                                                                                                  \n",
      " batch_normalization_1 (Bat  (None, 48, 48, 64)           256       ['re_lu_1[0][0]']             \n",
      " chNormalization)                                                                                 \n",
      "                                                                                                  \n",
      " global_average_pooling2d_1  (None, 64)                   0         ['batch_normalization_1[0][0]'\n",
      "  (GlobalAveragePooling2D)                                          ]                             \n",
      "                                                                                                  \n",
      " dense_2 (Dense)             (None, 4)                    256       ['global_average_pooling2d_1[0\n",
      "                                                                    ][0]']                        \n",
      "                                                                                                  \n",
      " dense_3 (Dense)             (None, 64)                   256       ['dense_2[0][0]']             \n",
      "                                                                                                  \n",
      " reshape_1 (Reshape)         (None, 1, 1, 64)             0         ['dense_3[0][0]']             \n",
      "                                                                                                  \n",
      " multiply_1 (Multiply)       (None, 48, 48, 64)           0         ['batch_normalization_1[0][0]'\n",
      "                                                                    , 'reshape_1[0][0]']          \n",
      "                                                                                                  \n",
      " max_pooling2d (MaxPooling2  (None, 24, 24, 64)           0         ['multiply_1[0][0]']          \n",
      " D)                                                                                               \n",
      "                                                                                                  \n",
      " dropout (Dropout)           (None, 24, 24, 64)           0         ['max_pooling2d[0][0]']       \n",
      "                                                                                                  \n",
      " conv2d_2 (Conv2D)           (None, 24, 24, 128)          73856     ['dropout[0][0]']             \n",
      "                                                                                                  \n",
      " re_lu_2 (ReLU)              (None, 24, 24, 128)          0         ['conv2d_2[0][0]']            \n",
      "                                                                                                  \n",
      " batch_normalization_2 (Bat  (None, 24, 24, 128)          512       ['re_lu_2[0][0]']             \n",
      " chNormalization)                                                                                 \n",
      "                                                                                                  \n",
      " global_average_pooling2d_2  (None, 128)                  0         ['batch_normalization_2[0][0]'\n",
      "  (GlobalAveragePooling2D)                                          ]                             \n",
      "                                                                                                  \n",
      " dense_4 (Dense)             (None, 8)                    1024      ['global_average_pooling2d_2[0\n",
      "                                                                    ][0]']                        \n",
      "                                                                                                  \n",
      " dense_5 (Dense)             (None, 128)                  1024      ['dense_4[0][0]']             \n",
      "                                                                                                  \n",
      " reshape_2 (Reshape)         (None, 1, 1, 128)            0         ['dense_5[0][0]']             \n",
      "                                                                                                  \n",
      " multiply_2 (Multiply)       (None, 24, 24, 128)          0         ['batch_normalization_2[0][0]'\n",
      "                                                                    , 'reshape_2[0][0]']          \n",
      "                                                                                                  \n",
      " max_pooling2d_1 (MaxPoolin  (None, 12, 12, 128)          0         ['multiply_2[0][0]']          \n",
      " g2D)                                                                                             \n",
      "                                                                                                  \n",
      " conv2d_3 (Conv2D)           (None, 12, 12, 128)          147584    ['max_pooling2d_1[0][0]']     \n",
      "                                                                                                  \n",
      " re_lu_3 (ReLU)              (None, 12, 12, 128)          0         ['conv2d_3[0][0]']            \n",
      "                                                                                                  \n",
      " batch_normalization_3 (Bat  (None, 12, 12, 128)          512       ['re_lu_3[0][0]']             \n",
      " chNormalization)                                                                                 \n",
      "                                                                                                  \n",
      " global_average_pooling2d_3  (None, 128)                  0         ['batch_normalization_3[0][0]'\n",
      "  (GlobalAveragePooling2D)                                          ]                             \n",
      "                                                                                                  \n",
      " dense_6 (Dense)             (None, 8)                    1024      ['global_average_pooling2d_3[0\n",
      "                                                                    ][0]']                        \n",
      "                                                                                                  \n",
      " dense_7 (Dense)             (None, 128)                  1024      ['dense_6[0][0]']             \n",
      "                                                                                                  \n",
      " reshape_3 (Reshape)         (None, 1, 1, 128)            0         ['dense_7[0][0]']             \n",
      "                                                                                                  \n",
      " multiply_3 (Multiply)       (None, 12, 12, 128)          0         ['batch_normalization_3[0][0]'\n",
      "                                                                    , 'reshape_3[0][0]']          \n",
      "                                                                                                  \n",
      " max_pooling2d_2 (MaxPoolin  (None, 6, 6, 128)            0         ['multiply_3[0][0]']          \n",
      " g2D)                                                                                             \n",
      "                                                                                                  \n",
      " dropout_1 (Dropout)         (None, 6, 6, 128)            0         ['max_pooling2d_2[0][0]']     \n",
      "                                                                                                  \n",
      " flatten (Flatten)           (None, 4608)                 0         ['dropout_1[0][0]']           \n",
      "                                                                                                  \n",
      " dense_8 (Dense)             (None, 1024)                 4719616   ['flatten[0][0]']             \n",
      "                                                                                                  \n",
      " batch_normalization_4 (Bat  (None, 1024)                 4096      ['dense_8[0][0]']             \n",
      " chNormalization)                                                                                 \n",
      "                                                                                                  \n",
      " dropout_2 (Dropout)         (None, 1024)                 0         ['batch_normalization_4[0][0]'\n",
      "                                                                    ]                             \n",
      "                                                                                                  \n",
      " dense_9 (Dense)             (None, 256)                  262400    ['dropout_2[0][0]']           \n",
      "                                                                                                  \n",
      " batch_normalization_5 (Bat  (None, 256)                  1024      ['dense_9[0][0]']             \n",
      " chNormalization)                                                                                 \n",
      "                                                                                                  \n",
      " dropout_3 (Dropout)         (None, 256)                  0         ['batch_normalization_5[0][0]'\n",
      "                                                                    ]                             \n",
      "                                                                                                  \n",
      " dense_10 (Dense)            (None, 7)                    1799      ['dropout_3[0][0]']           \n",
      "                                                                                                  \n",
      "==================================================================================================\n",
      "Total params: 5235335 (19.97 MB)\n",
      "Trainable params: 5232071 (19.96 MB)\n",
      "Non-trainable params: 3264 (12.75 KB)\n",
      "__________________________________________________________________________________________________\n"
     ]
    }
   ],
   "source": [
    "def squeeze_excite_block(input_tensor, ratio=16):\n",
    "    init = input_tensor\n",
    "    channel_axis = -1  \n",
    "    filters = init.shape[channel_axis]\n",
    "    \n",
    "    se = GlobalAveragePooling2D()(init)\n",
    "    se = Dense(filters // ratio, activation='relu', use_bias=False)(se)\n",
    "    se = Dense(filters, activation='sigmoid', use_bias=False)(se)\n",
    "    se = Reshape((1, 1, filters))(se)\n",
    "    x = Multiply()([init, se])\n",
    "    return x\n",
    "\n",
    "# Model definition, use Functional API instead of Sequential\n",
    "input_img = Input(shape=(48, 48, 1))\n",
    "\n",
    "# 1st conv block\n",
    "x = Conv2D(32, kernel_size=(3, 3), padding='same')(input_img)\n",
    "x = ReLU()(x)\n",
    "x = BatchNormalization()(x)\n",
    "x = squeeze_excite_block(x)  \n",
    "\n",
    "# 2nd conv block\n",
    "x = Conv2D(64, kernel_size=(3, 3), padding='same')(x)\n",
    "x = ReLU()(x)\n",
    "x = BatchNormalization()(x)\n",
    "x = squeeze_excite_block(x)  \n",
    "x = MaxPooling2D(pool_size=(2, 2))(x)\n",
    "x = Dropout(0.25)(x)\n",
    "\n",
    "# 3rd conv block\n",
    "x = Conv2D(128, kernel_size=(3, 3), padding='same')(x)\n",
    "x = ReLU()(x)\n",
    "x = BatchNormalization()(x)\n",
    "x = squeeze_excite_block(x)  \n",
    "x = MaxPooling2D(pool_size=(2, 2))(x)\n",
    "\n",
    "# 4th conv block\n",
    "x = Conv2D(128, kernel_size=(3, 3), padding='same')(x)\n",
    "x = ReLU()(x)\n",
    "x = BatchNormalization()(x)\n",
    "x = squeeze_excite_block(x)\n",
    "x = MaxPooling2D(pool_size=(2, 2))(x)\n",
    "x = Dropout(0.25)(x)\n",
    "\n",
    "# Flatten and Dense Layers\n",
    "x = Flatten()(x)\n",
    "x = Dense(1024, activation='relu')(x)\n",
    "x = BatchNormalization()(x)\n",
    "x = Dropout(0.5)(x)\n",
    "\n",
    "# Additional Dense layer to increase complexity\n",
    "x = Dense(256, activation='relu')(x)\n",
    "x = BatchNormalization()(x)\n",
    "x = Dropout(0.5)(x)\n",
    "\n",
    "# Output layer\n",
    "output = Dense(7, activation='softmax')(x)\n",
    "\n",
    "model = Model(inputs=input_img, outputs=output)\n",
    "\n",
    "\n",
    "# Set up the learning rate schedule and compile the model\n",
    "initial_learning_rate = 0.0001\n",
    "lr_schedule = ExponentialDecay(\n",
    "    initial_learning_rate,\n",
    "    decay_steps=100000,\n",
    "    decay_rate=0.96,\n",
    "    staircase=True)\n",
    "\n",
    "optimizer = Adam(learning_rate=lr_schedule)\n",
    "\n",
    "model.compile(optimizer=optimizer,\n",
    "              loss='categorical_crossentropy',\n",
    "              metrics=['accuracy'])\n",
    "\n",
    "model.summary()\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "873c9724",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/80\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "2024-03-22 01:55:02.485435: E tensorflow/core/grappler/optimizers/meta_optimizer.cc:954] layout failed: INVALID_ARGUMENT: Size of values 0 does not match size of permutation 4 @ fanin shape inmodel/dropout/dropout/SelectV2-2-TransposeNHWCToNCHW-LayoutOptimizer\n",
      "2024-03-22 01:55:03.508175: I tensorflow/compiler/xla/stream_executor/cuda/cuda_dnn.cc:432] Loaded cuDNN version 8600\n",
      "2024-03-22 01:55:04.371397: I tensorflow/compiler/xla/service/service.cc:168] XLA service 0x7f1215379e50 initialized for platform CUDA (this does not guarantee that XLA will be used). Devices:\n",
      "2024-03-22 01:55:04.371441: I tensorflow/compiler/xla/service/service.cc:176]   StreamExecutor device (0): NVIDIA GeForce GTX 1080 Ti, Compute Capability 6.1\n",
      "2024-03-22 01:55:04.379730: I tensorflow/compiler/mlir/tensorflow/utils/dump_mlir_util.cc:255] disabling MLIR crash reproducer, set env var `MLIR_CRASH_REPRODUCER_DIRECTORY` to enable.\n",
      "2024-03-22 01:55:04.596876: I ./tensorflow/compiler/jit/device_compiler.h:186] Compiled cluster using XLA!  This line is logged at most once for the lifetime of the process.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "717/717 [==============================] - 64s 75ms/step - loss: 2.7066 - accuracy: 0.2070 - val_loss: 1.9376 - val_accuracy: 0.3083\n",
      "Epoch 2/80\n",
      "717/717 [==============================] - 41s 57ms/step - loss: 2.3007 - accuracy: 0.2481 - val_loss: 1.6923 - val_accuracy: 0.3541\n",
      "Epoch 3/80\n",
      "717/717 [==============================] - 39s 54ms/step - loss: 2.0845 - accuracy: 0.2864 - val_loss: 1.6185 - val_accuracy: 0.3769\n",
      "Epoch 4/80\n",
      "717/717 [==============================] - 39s 55ms/step - loss: 1.9256 - accuracy: 0.3179 - val_loss: 1.5871 - val_accuracy: 0.3900\n",
      "Epoch 5/80\n",
      "717/717 [==============================] - 38s 53ms/step - loss: 1.7979 - accuracy: 0.3515 - val_loss: 1.5761 - val_accuracy: 0.4089\n",
      "Epoch 6/80\n",
      "717/717 [==============================] - 40s 55ms/step - loss: 1.7040 - accuracy: 0.3734 - val_loss: 1.5058 - val_accuracy: 0.4326\n",
      "Epoch 7/80\n",
      "717/717 [==============================] - 38s 53ms/step - loss: 1.6087 - accuracy: 0.4056 - val_loss: 1.5830 - val_accuracy: 0.4282\n",
      "Epoch 8/80\n",
      "717/717 [==============================] - 38s 54ms/step - loss: 1.5547 - accuracy: 0.4173 - val_loss: 1.4619 - val_accuracy: 0.4619\n",
      "Epoch 9/80\n",
      "717/717 [==============================] - 39s 54ms/step - loss: 1.4895 - accuracy: 0.4328 - val_loss: 1.3642 - val_accuracy: 0.4838\n",
      "Epoch 10/80\n",
      "717/717 [==============================] - 38s 53ms/step - loss: 1.4370 - accuracy: 0.4547 - val_loss: 1.4594 - val_accuracy: 0.4621\n",
      "Epoch 11/80\n",
      "717/717 [==============================] - 38s 53ms/step - loss: 1.4000 - accuracy: 0.4659 - val_loss: 1.3941 - val_accuracy: 0.4815\n",
      "Epoch 12/80\n",
      "717/717 [==============================] - 38s 53ms/step - loss: 1.3612 - accuracy: 0.4829 - val_loss: 1.3460 - val_accuracy: 0.4944\n",
      "Epoch 13/80\n",
      "717/717 [==============================] - 37s 52ms/step - loss: 1.3313 - accuracy: 0.4901 - val_loss: 1.2965 - val_accuracy: 0.5110\n",
      "Epoch 14/80\n",
      "717/717 [==============================] - 38s 54ms/step - loss: 1.3052 - accuracy: 0.5085 - val_loss: 1.2790 - val_accuracy: 0.5227\n",
      "Epoch 15/80\n",
      "717/717 [==============================] - 37s 52ms/step - loss: 1.2916 - accuracy: 0.5081 - val_loss: 1.2440 - val_accuracy: 0.5243\n",
      "Epoch 16/80\n",
      "717/717 [==============================] - 38s 53ms/step - loss: 1.2738 - accuracy: 0.5154 - val_loss: 1.2691 - val_accuracy: 0.5189\n",
      "Epoch 17/80\n",
      "717/717 [==============================] - 39s 54ms/step - loss: 1.2444 - accuracy: 0.5264 - val_loss: 1.2323 - val_accuracy: 0.5333\n",
      "Epoch 18/80\n",
      "717/717 [==============================] - 37s 51ms/step - loss: 1.2344 - accuracy: 0.5319 - val_loss: 1.2064 - val_accuracy: 0.5409\n",
      "Epoch 19/80\n",
      "717/717 [==============================] - 35s 49ms/step - loss: 1.2184 - accuracy: 0.5388 - val_loss: 1.1659 - val_accuracy: 0.5541\n",
      "Epoch 20/80\n",
      "717/717 [==============================] - 36s 50ms/step - loss: 1.2067 - accuracy: 0.5455 - val_loss: 1.1849 - val_accuracy: 0.5503\n",
      "Epoch 21/80\n",
      "717/717 [==============================] - 36s 50ms/step - loss: 1.1927 - accuracy: 0.5490 - val_loss: 1.1469 - val_accuracy: 0.5665\n",
      "Epoch 22/80\n",
      "717/717 [==============================] - 36s 49ms/step - loss: 1.1809 - accuracy: 0.5536 - val_loss: 1.1606 - val_accuracy: 0.5546\n",
      "Epoch 23/80\n",
      "717/717 [==============================] - 36s 50ms/step - loss: 1.1709 - accuracy: 0.5580 - val_loss: 1.1277 - val_accuracy: 0.5709\n",
      "Epoch 24/80\n",
      "717/717 [==============================] - 35s 49ms/step - loss: 1.1553 - accuracy: 0.5669 - val_loss: 1.1454 - val_accuracy: 0.5684\n",
      "Epoch 25/80\n",
      "717/717 [==============================] - 36s 50ms/step - loss: 1.1413 - accuracy: 0.5722 - val_loss: 1.1224 - val_accuracy: 0.5814\n",
      "Epoch 26/80\n",
      "717/717 [==============================] - 36s 50ms/step - loss: 1.1410 - accuracy: 0.5719 - val_loss: 1.1463 - val_accuracy: 0.5677\n",
      "Epoch 27/80\n",
      "717/717 [==============================] - 36s 51ms/step - loss: 1.1260 - accuracy: 0.5747 - val_loss: 1.1028 - val_accuracy: 0.5761\n",
      "Epoch 28/80\n",
      "717/717 [==============================] - 37s 51ms/step - loss: 1.1163 - accuracy: 0.5819 - val_loss: 1.1387 - val_accuracy: 0.5716\n",
      "Epoch 29/80\n",
      "717/717 [==============================] - 36s 50ms/step - loss: 1.1178 - accuracy: 0.5733 - val_loss: 1.1101 - val_accuracy: 0.5798\n",
      "Epoch 30/80\n",
      "717/717 [==============================] - 35s 49ms/step - loss: 1.1000 - accuracy: 0.5882 - val_loss: 1.0843 - val_accuracy: 0.6007\n",
      "Epoch 31/80\n",
      "717/717 [==============================] - 36s 50ms/step - loss: 1.0953 - accuracy: 0.5886 - val_loss: 1.0943 - val_accuracy: 0.5821\n",
      "Epoch 32/80\n",
      "717/717 [==============================] - 36s 50ms/step - loss: 1.0912 - accuracy: 0.5889 - val_loss: 1.0680 - val_accuracy: 0.6044\n",
      "Epoch 33/80\n",
      "717/717 [==============================] - 36s 49ms/step - loss: 1.0840 - accuracy: 0.5935 - val_loss: 1.0566 - val_accuracy: 0.5967\n",
      "Epoch 34/80\n",
      "717/717 [==============================] - 36s 50ms/step - loss: 1.0734 - accuracy: 0.5942 - val_loss: 1.0676 - val_accuracy: 0.5948\n",
      "Epoch 35/80\n",
      "717/717 [==============================] - 36s 50ms/step - loss: 1.0667 - accuracy: 0.6016 - val_loss: 1.0622 - val_accuracy: 0.5965\n",
      "Epoch 36/80\n",
      "717/717 [==============================] - 37s 51ms/step - loss: 1.0599 - accuracy: 0.6045 - val_loss: 1.0628 - val_accuracy: 0.5960\n",
      "Epoch 37/80\n",
      "717/717 [==============================] - 36s 50ms/step - loss: 1.0486 - accuracy: 0.6057 - val_loss: 1.0581 - val_accuracy: 0.6021\n",
      "Epoch 38/80\n",
      "717/717 [==============================] - 37s 52ms/step - loss: 1.0534 - accuracy: 0.6022 - val_loss: 1.0388 - val_accuracy: 0.6117\n",
      "Epoch 39/80\n",
      "717/717 [==============================] - 37s 51ms/step - loss: 1.0454 - accuracy: 0.6073 - val_loss: 1.0452 - val_accuracy: 0.6063\n",
      "Epoch 40/80\n",
      "717/717 [==============================] - 39s 54ms/step - loss: 1.0414 - accuracy: 0.6093 - val_loss: 1.0242 - val_accuracy: 0.6063\n",
      "Epoch 41/80\n",
      "717/717 [==============================] - 36s 50ms/step - loss: 1.0296 - accuracy: 0.6148 - val_loss: 1.0367 - val_accuracy: 0.6053\n",
      "Epoch 42/80\n",
      "717/717 [==============================] - 37s 51ms/step - loss: 1.0343 - accuracy: 0.6112 - val_loss: 1.0295 - val_accuracy: 0.6112\n",
      "Epoch 43/80\n",
      "717/717 [==============================] - 37s 51ms/step - loss: 1.0219 - accuracy: 0.6170 - val_loss: 1.0398 - val_accuracy: 0.6074\n",
      "Epoch 44/80\n",
      "717/717 [==============================] - 37s 51ms/step - loss: 1.0161 - accuracy: 0.6167 - val_loss: 1.0259 - val_accuracy: 0.6128\n",
      "Epoch 45/80\n",
      "717/717 [==============================] - ETA: 0s - loss: 1.0126 - accuracy: 0.6184Restoring model weights from the end of the best epoch: 40.\n",
      "717/717 [==============================] - 37s 52ms/step - loss: 1.0126 - accuracy: 0.6184 - val_loss: 1.0462 - val_accuracy: 0.6072\n",
      "Epoch 45: early stopping\n"
     ]
    }
   ],
   "source": [
    "# Early stopping\n",
    "early_stopping_callback = EarlyStopping(monitor='val_loss', patience=5, verbose=1, mode='min', restore_best_weights=True)\n",
    "\n",
    "model_info = model.fit(\n",
    "        train_generator,\n",
    "        steps_per_epoch=22968 // 32, # number of training data divided by batch size\n",
    "        epochs=80,\n",
    "        validation_data=validation_generator,\n",
    "        validation_steps=5741 // 32,\n",
    "        callbacks=[tensorboard_callback, early_stopping_callback]\n",
    ")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "aca431d6",
   "metadata": {},
   "source": [
    "- early stopping at 52/80, best epoch 47, accuracy 63%"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "53362a2e",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x360 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "accuracy = model_info.history['accuracy']\n",
    "val_accuracy = model_info.history['val_accuracy']\n",
    "loss = model_info.history['loss']\n",
    "val_loss = model_info.history['val_loss']\n",
    "\n",
    "plt.figure(figsize=(12, 5))\n",
    "\n",
    "# accuracy graph\n",
    "plt.subplot(1, 2, 1)\n",
    "plt.plot(accuracy, label='accuracy')\n",
    "plt.plot(val_accuracy, label='val accuracy')\n",
    "plt.title('Accuracy for New SE Model on FER-2013')\n",
    "plt.xlabel('Epochs')\n",
    "plt.ylabel('Accuracy')\n",
    "plt.legend()\n",
    "\n",
    "# loss graph\n",
    "plt.subplot(1, 2, 2)\n",
    "plt.plot(loss, label='loss')\n",
    "plt.plot(val_loss, label='val loss')\n",
    "plt.title('Loss for New SE Model on FER-2013')\n",
    "plt.xlabel('Epochs')\n",
    "plt.ylabel('Loss')\n",
    "plt.legend()\n",
    "\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "74c2e5e7",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "225/225 [==============================] - 11s 48ms/step - loss: 1.0031 - accuracy: 0.6237\n",
      "Test loss: 1.0031490325927734\n",
      "Test accuracy: 0.623711347579956\n"
     ]
    }
   ],
   "source": [
    "results = model.evaluate(test_generator, steps=len(test_generator))\n",
    "\n",
    "# Test results\n",
    "print(\"Test loss:\", results[0])\n",
    "print(\"Test accuracy:\", results[1])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "4b7149f8",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1/1 [==============================] - 0s 28ms/step\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAWsAAAEYCAYAAACa8NcXAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjQuMiwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy8rg+JYAAAACXBIWXMAAAsTAAALEwEAmpwYAADcvElEQVR4nOz9aZBk2XUeCH7X93332CMzcqmsqqwNha0ggAUUFwwpihxSEM3EJsdImoY0qU0kmz0jtkQNx4xqUepu9tioWxuFAVukRC0D4yKSICkNQHEDiL1QqC2rsnKJzFg9PHzf9zc/PL4T572MrMxEeWRmwP2YhUWEu7/F373n3HO+851zjWVZmMlMZjKTmTzc4nrQNzCTmcxkJjO5s8yM9UxmMpOZnACZGeuZzGQmMzkBMjPWM5nJTGZyAmRmrGcyk5nM5ATIzFjPZCYzmckJkAdqrI0xv2aM+cUHeQ8zebAymwMzmc2Bu5M7GmtjzA1jTNsY0zDG7BljftUYE7kfN/c29/SrxhjLGHP+Lj77/MG9N4wxzYPjGurn1P2454N7+QVjTN9x/bP36/rfqMzmwOTFGOMzxrxpjNm639f+RuRhmQPGmBeMMSPH+P3oXRz30MwBY8zPGmNeM8bUjTHrxpifvZvj7taz/l7LsiIA3g3gfQB+/ogb8Nz97X7jYoz5FgDn7vbzlmV91rKsyMH9P3HwcoKvWZa1oc59P77DJ9W1I5ZlXb8P15yEzObAZOVnAeTv07UmJQ/LHNhx6NC/udMBD9kcMAB+BEASwHcB+EljzA/e6aB7gkEsy9oG8J8BPAkAB6vT3zbGXAFw5eC17zHGfN0YUzHGfN4Y87TcoTHPGmO+drCifBJA4F6uf/AQ/xmAn7yX497mfL9gjPlNY8y/M8bUAPyYMyQ7WMm31P9LxpjfMsbsH6yKPz2JezkpMpsD73wOGGPOAPi/APifJvEd7rc86Dkwabnfc8CyrF+yLOtrlmUNLMu6DOB3AXzoTsfdk7E2xqwC+G4AL6mXvx/AcwAuGmPeDeBfA/ibANIAPg7g94wxfmOMD8DvAPh1ACkAvwHgrznOXznwmm4n/z2AP7cs65V7ue87yPcB+E0ACQD//u0+aIxxAfgUgJcBLAP4dgA/Y4z5zoP3v8UYU7nD9b7XGFMyxrxujPlv3+G933eZzYGJzIF/BuDvA2i/o7t+QPIQzIE5M4Zi1o0x/8QYE57A17rfc4DnMgCeB/D6HT9sWdbb/gC4AaABoALgJoB/CSB48J4F4NvUZ38ZwD90HH8ZwEcAfBjADgCj3vs8gF+80z0cfHYVwFUAcXXt83dzrDrH2sFxnoP/fwFjxdef+TV9TwBeALB18PdzADYcn/85AL96l9e/CGAJgBvABwHsAvhv7uU7PIif2RyY6Bz4qwD+i/O8D/vPQzQHFg70yAXgDIA/B/DxkzQHHMf9A4yNvv9On71bbOb7Lcv6o9u8t6n+Pg3gR40xP6Ve82FsoCwA29bBHR7Izbu8PgD8bwD+R8uyqvdwzN3I5p0/InIawJJj1XQD+OzdHGxZ1iX17+eNMf87gB8A8B/v4R4elMzmwFi+4Tlw4AH+EsZe6UmUBz4HLMvKAcgd/LtujPkfAPwBxl78O5H7ZgcoxpifxBi7ft6yrO6dPj8JIF0/9E0A/8iyrH90xI19BMCyMcaogToF4NpdXufbAXyLMeaX1GtfMMb8d5Zl/Ydv5MYPxNl2sAkgpP5fUH9vAli3LOuRd3A957XNhM71IGU2B+5OHsHYq/vsOPqFD0DcGJMD8AHLsm58A+d8WOR+zYGjrjsJHbqvdsAY8zcA/D0AH7Ys664YQZPmWX8CwN8yxjxnxhI2xvwVY0wUwBcADAD8tDHGY4z5GID338O5LwB4BsC7Dn4A4HsB/CdAkgR/OoHv8HUA322MSRljFgD8jHrvywBqxpi/a4wJGmPcxpgnjTHvu5sTG2O+zxiTPHg27wfw0xgnF76ZZDYHbi+vYQzl8P5/HMDewd/34tk97HJsc+Ag0Xfq4LyrAP5nKB06AXMAxpgfBvCPAXzUugc22ESNtWVZXwXwEwD+OYAyxvjijx281wPwsYP/ywD+OoDf1sebMd/x+ducO29ZVo4/By8XLMtikmYVwF9M4Gv8OsYY0g0AnwbwSXUPQ4yNw7sArAMoAPgVAPGD+3/eGNN4m3P/IMbPpA7g3wL4X6y7oB2dJJnNgdvPAWuc/df3XwIwOvh/OIH7fijkOOcAxrTBL2Ds+X4e4wVQMzEe6jlwIL+IceL1K+aQ5/2v7nRDxg4dnVwxxnwdwLdbllV80Pcykwcjszkwk2/mOfBNY6xnMpOZzOSbWWaNnGYyk5nM5ATIzFjPZCYzmckJkJmxnslMZjKTEyATb1jy/d///ZYxBsYYuN1ueDwe7O7uYnNzE9evX4fH44FlWRgOhwgGg3C5xuvFYDCAy+WC1+uFx+PBcDiUv91uN1qtFvr9PizLgsvlgtvthmVZGI1GGA6HGA7HyXRjjPwNjCs0k8kkRqMR+v0+BoMB/H4/QqEQotEoLl++jEcffRTf8i3fgr/zd/4OGo0GGo0GKpUKrl69itXVVSwtLWF1dRVXr17F3t4eqtUqWq0Wzp49C7/fD6/Xi7m5Ofke/X4fH/jAB74Z+NPfqFiWZaHT6WBnZwc/93M/h83NTbRaLXi9Xug8yWg0wmAwwGAwwGg0wgH/WObPwckwGo0QDofh9XoBAJ1OB4PBAG63G263W85zxI3Yfuvr8nXOF87F0Whke304HMo9eL1emW+WZcHtdqPT6SAQCOCnfuqn8EM/9ENYWloaV5x5PFM7B5566inrO77jO/C+970Pzz77LLrdLqLRKOLxOEKhENxuN4wxGI1GcLvd8jxpO4bDoeiSMQb9fh+tVgudTgehUAh+vx8ejwcvvvgi+v0+fD4fFhYW0Ol0ZHyazSZarRaGwyECgQC8Xi9arRbK5TK++MUv4urVq3C5XMhkMhiNRmg0Guh0Ouj3++h0OjKncrkc6vU6BoOB3PPdSr/fn9gcmLixHgwGolDAWOmCwSBisRg8Ho8opM/ng8/nEwNqWRa8Xq88DA4SlXY0GsnPcDhEv98HABlgRwmnGPPBYIBer2c7V7fbtRn9er2O/f19tFot+P1++Q6DwQBvvfUWWq0WEokE5ufnEQ6H0Wq1UKvV4PP5AIyVuVKpAICcc5qFY8zFl8+Dk5zGkAZRG0sqMQD5m5/l30dd73aJ8qOOuZ2hdioh5xANtcvlks/zPd4rlZ3z6l4U+ptRnnnmGTz77LO4cOECPB4P/H4/wuEwAoFxzyY9BtpIc+44x4hGXes5jTgXbX0+l8sFn8+HbreLTqeDQqGAeDyOfr+PRqOBQqGAUCgkY+r1euXe6AAMBgN0u134fD6xQfdqrCcpEzfWVBxOYgDw+/2IRCLw+/1ot9tirD0eD3q9nqyqVAoOkPM1598AxCjo9/g/cDjY9NA5wFypPR4Pms0mCoUC9vf3cfr0afj9fgQCAYRCIVy/fh0ulwt7e3tYXV1FKpVCJBJBOBxGo9GQa9ZqNQCQhWmahePvcrng9/ttc4HGk56TFkZi/LzL5ZK/j/KQtWF/O1aTvr7+HP8+yrg675mfcc5tCo31tBtpyrPPPovz589jYWEBg8EAwWAQgUAAHo9HjCt1FoCMu35+Ry3odNgAiC5Tvyk0/IziaaxDoRD6/b4YcN6PMQZ+v992bbfbLbaJ8/Cocb+fMnFjTUNpWRb6/T7cbjd8Ph/i8ThisRh6vR7cbjfC4bDNU+aD52sMSekdO5VHK6keQB5HyISD5vF4ROH8fj+GwyFarRYCgQBarRY2Nzfx5S9/GUtLSwiFQjDG4LHHHsPOzg7y+Ty+9KUvweVyIZvNIhQKIRQKyb2MRiNcuXJFwuJEIjHpx3qihGPgcrkQj8dlYaYx7nQ6Em7G43GbcvFYwO590chrhXIqj1Yqp/HmsTrM5udvB5XoaEwbaUI2zmiwUqnIvJt2+Z7v+R7EYjH4fD5YliWLNm2Cdqj0uACHY6VhqMFggOFwiF6vh0AgIJ8h7EF4jQaa52s2mygWi6jVakgkEqjVahiNRvD5fAiFQmKsA4GAOIha2u22RPG81wclx9pkmwNCnKhWqwl+FI1G0ev10Gw25fN6sDg4Xq9X4AYOgnP11YPNsIZKTwMQDocRDAbRarUAQDxrYLyKFotFfPzjH8f8/DyefvppLC4uwu1247nnnsPu7i5yuRy+9rWvYWlpCXNzc1hcXEQ8HpeV/5FHHkE4HBaM68KFC8f5aE+EtFotXLp0Cfl8Xryrer0OAAgEAhJtUbTxpdeto6heryfvawPrhEE4JtpDpxzluenXdcTGY/UxGuI5SrQXPs2yuLiIbreL4XAIn89ne2Y6N0FHS8OaNNQ0vjTWjMr9fj/8fr8NygwGg4hEIuIktlot5PN5lEolDIdDnDt3DrFYDNVqFdXquA9YOBwWj7pSqYiTSAeu2WyiUqmg2Wwe6TjcbzkWGITC1a7b7aLRaKDX6yEYDCKVSmF5eRnlcllWxm63Kx40MSOdYHy7kNcZ5mqFomcfDocRi8WQSCQQDAbR7/dRLpdRLpcBQP7/vd/7PZTLZbzvfe9DMplEKpWSyVCr1VAsFtFutzEYDBCPx22JknK5LHj2NAsnfa/Xw/b2NprNJgaDgXgxjHR0tEM5CgLj606j6fyM8xz0nN/OcNLL1oZbG2zn6/o9/QNAklm3g0qmSYgxAzgyh3O7xVWPF20BX6MDx2MJgdBOeL1e9Ho99Ho91Go1NJtNeDwexONxLCwsSL6KDmIwGITX65UoLxAICFbd6/UkAiRmDhwu3ncy2MdhzI/NWOuH3uv1ZCJHIhFks1mcOnVKMGvN6KCRZgJSs0M03qnxTK0YxEm9Xi+8Xq8sDul0GolEApFIBKlUCr1eD7lcDm+99RaAcbjT6/Xwmc98RqCa97znPfB4PEgmk4hEItjc3MT+/j7q9TparRbm5uYQDAYlxKtWqzLwMxmHr9VqFb1eT9g+HBsaau1BM+x1/u2EuoBDj8yZu9BKdSec8ajEtDb+ToV0et38m5/jd9Uw3rQKF2fN1uFzOWqB1qKdLe2BEyblGDmNtc/nQ7vdRrvdFl30+/0IBoPIZrPindfrdYFm3G63GG/am2azKYa60+nYHECd5LzfHvbEjbUOd8jaaLVaaLfbGI1GSKfTOH36NC5cuID5+XnMzc0hl8thfX0du7u7AMaKxpWSxhsYey5cYbni8oH5fD6BVy5cuIC5uTmk02ksLy8jm83KoEciEczPz6Pb7WJnZwd//Md/jCtXrqBQKKDdbmNnZwcvvvgiCoUCLMuSRGMsFsPp06cRDoexs7ODr371q2g2m0gmk5ifn8fp06exurqKU6dOTX2Ckc86EAjgkUceQTKZxP7+PkajETKZDHq9niR6AIjnchRLw8nGcF6H1yKVi06ChkyAW2EVZzLJiYU7P6OVUntamimwtbWFarWKbreLQCAw1d619qZ1cpb0R463Mzrhs9eLL71lPYYAbLaATlOn00G9Xkej0ZCkYyAQQDqdxhtvvIH9/X1UKhU5N3NX6XRanEdCH6TyaafCia3fT4M9cWNNyh0NbbfbRa/XE8VcXl7G2bNnce7cOdTrdSwtLQk2VCqVZDUrl8uCWRGzbjabooSRSEQyzOFwWDznTCaDubk5JJNJhMNh8dAZQjG5QTzt6aefRigUwtbWFq5duya49s7ODkqlEh599FG4XC7s7Oxgc3MTzWYT3W4Xc3NzNjri17/+dVy9elVCsF/5lV+Z9KM9ceLz+bC4uIhIJCK4X6/XQ7vdFqPZ6/VsnooTK9bKTIwSOAyz6alxcafoRLeTJsbzam6+pn7xt/a8dVTHY/SiYIwRKG/mWR961hxjAOJpayKBXog5zprPrOmx3W4Xfr9fnm273Uaz2UQsFpPols+fdsPr9SISicDtdmNvbw+FQgGdTgenT5+W8fT7/fD5fKhUKiiXy8jlcrLo0m5Q9N/A8cAdt5NjSzBykne7XfkBgFgshlgshnA4DLfbjWg0KlneTqcjtJp8Pi9wBzPuzLR7vV7hbDJRRagiFoshGo3aCii4MnJ15IAyPGq1WhgMBtjf34fH40G/30ez2cRrr72GlZUVZDIZZLNZFItFBINB+Hw+JBIJxGIxdLtdlEolvPzyy4KZMYEx7cJkEJVUh7JaUZ3G7XaJQe2pas+Nhlcb3btJ8mmP76h710KDrb1wpzFnEYdToadZNHTEsdbP3LmA8jkCkOQk/2ZBjDbWLF6hjeC16G2HQiEEg0H0ej0Ui0Vhk8RiMbkui3KYVGy1WuLMaW69HuvbfcfjlIkba2cBBDGkbrcLYwzC4TDC4bDQZRhG8ofHVqtVm1JwAEhepyfl8XgkUUAvSxtpTSMjx5Pn5KqbzWYFw7569aokGT7zmc/giSeewNLSEs6fP4+trS1JOHDQCoUC6vU6vvCFLwgOO83hL3BrYo7hKBdJ5iecOLQOfZ24sKZ6AYcejvag9YJMr07fh/Mencbh7b4PABs7gfetqxsHg4Eo+rSLc/HlcyZsoT1pJ3GAjA4ANoZHs9nE/Py8GHAmEW9nrEOhEBKJBLxeLxqNBvL5vEBx0WhUYFWPx4NGo2GDa3l9Om+8Lxr3280r/Z0nLceSYKTCuVwuAesJQdAjpoccDAbly+vQJZPJSFJSh0msTAIgnjVZJ9qTcyaN6JlzFeWABgIBBINBRKNRzM/PI5FIwBgjC8wnPvEJvPTSS/jBH/xBvPvd77bBKW63G9vb23jzzTeRzWaRyWTgcrnQ6XQm/VhPnHAOFItFlEolNJtN8Y663S7cbrckeI7ypAG7kdSMAc4Bsn3ILNEhKxd+Zw6FNDD94/SWnV4/5x8hL84rZ9uDSqWCnZ0dFItFaT8wrcLKP2fy3+v12rjoTgjMGCOwJZO1jFj6/T4ikQgCgYAUujDKZt3G1tYW2u22FK/xXPl8Ht1uF+FwWKJ5OnXD4VAopYyO6XDoCmt+h9vRP4Hj9a6PxbPW1JpOpyO4pM/nk7LTZDIJn88nxphet1ZYnTDSCu12u1Gr1SSh43K5pF8AQxwOVjAYlAGnEebDJuuE95ROpxGLxSS8CoVCaLVaePXVV4V+eO7cOaTTafh8Prz55pu4dOkSrl+/Lvi1xlWnVah8VAL2dOBzp6fd6/XE0AKHCSOKNuTGjNsW6Pe08dX4Nz9P0V4cj+N96HyGbmOg2Qic087+FfyuXLiHwyGKxeIMBsOhUev3+2KkGVXpyMuJV/N50lD2+31hdjEa5lhTR0OhkDBBms0mXC4XUqmUGPler4dGo4FIJGKL6LrdrrBDdKTGRCUjdwBSzaj5+/z8UQbbmaCehEzcWDspUMTw6EnTIxqNRqjX6yiXy2i323C5XDh//rx42ppHe1S4sbe3h6tXr2JzcxNerxfRaFQwqng8jsXFRaTTaWQyGVtm3hle8+ETEmFVEzBmGPT7fZRKJbz00ktYWFhApVLBysoKEokEvvzlL+Pq1auo1WrIZDLieU17bxDgcPyZUOz1erZCJcIWGq900uYoNNo8Tl+Dyke2wFGYIo0BPSmNjxIq08ZC3782yk5GAOem9txnMMih6OesjbN26G4HhfD/brcLr9crUZiObJgcZHRN6iTZW/l8XmAry7JEn3ksWUmDwUD6gpAgkclkxIvP5/PI5/PieGjGir7X22Hak5KJG+teryeroPaSiFH7/X50u11cuXIFly5dwpUrVyRE+fEf/3E8/vjjmJ+fFwV30mPoWX/pS1/CV7/6VVy/fl08XUIalmUhHo8jlUrh9OnTeM973oNsNotYLCYrp2VZaDQaKJVKtuRnOBwWo8vMM1fU3/iN38Af/MEfIBKJIB6PY3NzE+FwGKlUSiAZTTGaVuFY6eICZvbpTVP0PCHUoJWApeosU9eFEVRcetcartD3oZVJezuDwUC4uMyBhMNhW6Kb8Jy+16OMCwBbD5rjVNqTIGRu6OSy5k9r3rJerDmWHo8HkUgEnU4HmUwGwWAQHo8H9XodqVRKIBYaa7/fj1qtJsdGo1HcvHlToh521qOXT4gVAEKhELLZLPL5vIznd3zHd+DJJ5/E8vIyXn31VXz605/GlStXsLW1dUuCVM9XvchP2mk7Fs+aNz8YDFCv19HtdgVnfPPNN7G+vo5KpSK0u7Nnz2IwGOD69etiRAOBgDAHqJwMU8mhfNe73oUnnngCfr8f8XhcVthoNCohVK/Xw5e//GV0u12MRiNks1mcP38eiURCjDsfvsvlwsrKCorFIgqFgnTnYxKToVC1WkWj0ZAERjweF8zUmRybRqFC+v1+zM/PIx6Pi5FzGmrdA0YnJGk86TVXq1WbMfT5fGIAdLjKc/H8AGzYqcaiubDu7+9LhMWwmnOTfHBeW2PjOsnI683NzSGVSh3zE374RUcshDS4yGkdoX4DkPmhHSQNbY1GI+RyOXHO4vE4NjY2bOwN3To3FotJG2Y6C4zwC4WCNJjz+XzIZDKoVquIRCL40Ic+hB/+4R/GaDTC3t6enFs3H9MUUp1L4ffj95mkHEtRDL8AH5z2ioExvMAwg4rByjYyLPi/BvnJ1+YKSnyL8AoVLBqNAjisciuXyxLCcPVkIUs6nYbf70er1UKj0RBPTofGeuLxOzI0o5HgKupsOjWNQhzP6/Uim80im82i3W6j1WrdwmfWXgo9aUJlnU5HChNYVaqrWjlftEfORZ3n4+u8Lhf8Wq0mY67zGHQQOCcZyelzH1XwMRqNMD8/j1OnTiGTyUz9HND5Bqcx09DIUTRHTQBgQQzhVCarPR6PLKaNRkNsAmG3drstDdmYC2FiOxQKYW5uTsbS5/NJEYwxBhcvXoTL5UKpVML29jb29vYErqXNoTgdDr6m/5+UHAvPmquQprkBh8oTi8WkzSh/EokE+v0+qtUqyuUyjDGIx+NwuVySqOQP+4xQsVwuF9rttnAr2SyKyUOyTiqVCjY3N3Hz5k0MBgNkMhksLS2JQrNaiT86GaXZKGz2Qo9NJ02mPblI4aROpVLIZrPY39+XiIjRF/nvVF6Px4NAICAGkL1YyApgr3ENW5D6qXMhnH98TQvnSqPRQL1el1JjKhiVnZ8lJRQ4XFjIQOl2u1L1ZlkWstmseGrTDoM42RZOWiuNqObY66iHxp5wFGs1KpUKWq2WsMoGgwFqtRr29/dtlbFkd2ib0W63ZWyWlpZQKpWkDUI+n5do/+zZs2g0Gsjlctja2hJjrcvX9QLiTDLqqGGSciwVjAxRnV+IA+hyjZv8sGsdveRUKoVut4v9/X0UCgUMBgNJ+HW7XVSrVaHUUSFZ2VQqlWy9QOi1sxETDYFufl4ulxGPx6Uq0em50YD7fD5bOSszyWSzcEWfVa8dCo0VoSINN/F94DDa8ng8SCQS8Hg8qFQqKBQK2N7etuF/9KLcbjeq1SrS6bS03g2FQhKJEabQ1D6WEReLRWxuboqnHo1GEY1GZZ5p+p9lWTIv6J3pnW5oLOiYVCoV/Nt/+2+xtbWFj33sY4hEIg/gyT8csr6+jrW1NcTjcZvR4gJNWMTtdoujBYztBzvhdTodbG9vw+/3o9lsolQqIZ/Po1qtCn+62+1ifX0dtVoNCwsLspPP7u4uNjY2pPEaI3ZgXPvx3ve+VxKHjUYD29vbqNVqWFxchN/vR71eR71el6ZtXCCeeeYZnDlzBpcvX8bGxgZ2dnZs0QI55BremZQci2dN2ILKyS/CB+73+xGLxbC1tYVAIIBEIoFUKiXhS7vdRqVSESofcSe9ApOrSVYJPR4ODBNbOqFFvEl38SuVStL0hd4RaYWAPWngxKM1V7fT6dhw02kWRhjD4VB21NG8Z36GEQmz7lQo8tQXFxdtSqBD2l6vh3q9LsUSLpcLyWRSFlkNXxB7rFaraDabNuNMIcTGHIU29E58UsMixWIRLpcL0WgUiUQCN2/exNWrV1EsFqfaWBN2IPWOETb1hc+UY8mxYG5pa2sLm5ubKBQKiEajKJfL2NnZkUVXjzW3DHzuuecE5trf38crr7yCGzduoFqtCn/a5XKhXq/j2rVrqFar2N/fx8bGBnK5nMxFQiiM6jc3N6XFxNmzZ21bDDoTqKzlAI7uNvhO5FgTjE7iOzvSsQufLjEnN5LeM7EqJpB0ckl7S+RK0osj44BhECEXhsTsIcCBJg4FjFdc9r0meZ/3ryeY/k405Pr9SYc/J1HY4SyfzwtkpEXT6aiohLUAyOedXpnOUWg4hdEWObkaG9dbu1EJuQEF8yqascIwnEVTnM800vSeOG8YkQFApVKRUH2aJR6PS34pEonYWj5YloVgMCgEBBo68vL39/fx1ltvYX19XZqzVatV5HI5BAIBecaEw+r1OnK5HBYWFoSN1mg0sLm5KbUYHCPaCPa1LpfLAoHQkLdaLblnEh44V4rFosBnGv4E7JTO44DAjq2CUQu9KHpMNLKBQEAgjCtXruDixYsS6moCOndlYfKJJaI6mcCeIysrK8jn87h27RoqlQrS6TQWFhakXwg78NFj0n1s9/f3BQuj0efDHwwG8ppOYPE764KOaTfWxBorlQquXbsmSWQdIrIgJRQKodPpSLjJBC8NKCM0reSMxmhwycyg98aFmfOMjYTY8IdzkREcnQbya0nlJKNAi25lQMOt23O6XC7pMTHNsri4iJ2dHVQqFczPzyMYDIpj5fV6EQqF0O12sbu7K/Bnp9NBLpfDn/zJn+DSpUvY2dnB448/Lkb9xo0bWFxcxP7+PkKhEJLJpNDwNjY2ZDMBRlM3b95EoVAQe0KYjW1RC4UCCoWCra8+mV7z8/OyOBPqKhQKeOmll2RuseCHkM5RVa6TlIkbayc1y1nZc+XKFYE9fv7nfx6vvvoqvvSlL+G3fuu3sLOzY6N6pdNprK6uYnV1FaVSSUqU2a+WRvKZZ57Be9/7Xly4cAHnz5/Hv/gX/wIvvvgiNjc3US6X4Xa7sbm5iU6ng0QigXPnzuHs2bPC6KCip1IpyTTHYjFRYF38oMNyGgpN8p/JWNhv4TOf+Qy8Xi8SiQQef/xxXL9+HZFIRHIX3GaJSVsu1kwih0IhDAYDFItFMcTkwlPpfT4fWq2W7BYSjUaF4VOv11GpVBCNRsXjLpfLsuUUw+hIJCJzg/glI0Duxq2LOrhQa2guFArJAnDlyhW8973vfZBD8EBlf38fm5ub4hGvrq6Knq2srGBvb086bbpc466W29vbuHbtGv7sz/4Me3t7aDab6Pf72N7exmAwEAj1+vXr6HQ6WF1dlWhrNBrh+vXrAA6jOtZQ6MK3dDqNubk5dDod7O3tSWLy7NmzqNfrCAQCKJfLEinTQSQywNe1vuu5oROok5Zj2d1cFyVo8J2JOSrAzs4O3G43lpaW8Oyzz+L06dMCWbRaLdk0IJVKyWpF5WVXPgACk7CXbTwex/nz52WVPX36tChuv99HNptFKpWCx+ORz7TbbRSLRYFqXC4XEomETBgmrDRVh9+Rk1BXyE27cNOG9773vbao5FOf+pSMg+Yq8xmyNL3b7SIYDCKTydgMYTQaFcPaaDTQ7/fFcJMumEqlEAqFUK1WbWOnm9TTM1tZWZHGYjTmnBfAYaWbswUrcLjBL/+ORCLCbpn2DShY51Cv17G7uysLNhszsa90o9HAYDDAG2+8gZs3b2J7e1teAyB0OgqNLNlBZIDpNhGdTkdIA5xfhGD4/+bmpiwI5Ftb1rh/PfddpEft5FDrIh79AxxvB75jzYg54QIA4kEFg0FsbGxIEctTTz2FWCxmI5Uz0x8MBmV/Q60wuvyXWBcH79SpU9LcZWFhQTZB6HQ6WFxcRCqVsu1awgeuYY5IJCLUPGLmbyeaOzztQuP16KOPotFoyFh87nOfs/Ww1o1xtELRuFIhvV6vGGuGu1Rk3bGRRiEWi0lYSyyaWDfbZpI+ykQY74nX1fxs4NZ5rOmdzK3oOTXNwmfJTT7m5+elvzz71XMxJRtjY2MDe3t7kuwnG4uwJUvH6XgR/mQjtkajIVCLzhloh4ostc3NTdmNng4iF2saet3uVhf4PCha5rH2s9ZNU+hdZ7NZnDlzBmtra3jjjTcQj8eRTCbx3HPPoVqtSnImk8kgnU4L0B8MBoX/SipdKBRCJBLBjRs3sLGxgUKhgM3NTcRiMZw/fx7nz5+XBAQTFaxcZJkqPTN6f7xfAFK6Tm9JF2HolZTH6EkxzcLJHAgEsLq6KhvmVqtVaUtJmmW73ZbFkMUpHFseEw6HpWc5pd1uIx6PIxgM2qrU/H6/LdTd398HAKlM5MLPBHO320UikRCPrN1uS1KRPH1+lkaD85nJTHrfvH4sFkM6nX5Qj/+hEC6gZOIUCgXMzc3BGIPd3V0UCgW0Wi1YloVKpYK9vT3s7e0hn8+LsXa73WI02aANsLdAJTU0kUgIa0PvEqSjNyaii8UiXnvtNSlpj8fj8v5gMMDe3p7wsnVjL8C+sbIzyj5uORZjrWv8na+vrKzg4sWLeOKJJ6RjHgDBfokPGmMkaUgF0ed2uVzynq71J61Pe1GZTMaW3ec1+/2+DJTX65XdI0qlkmDkTHgBR2/Sqo01ZdqNNUXjd6VSSWhUTDYC42pWKhcNKY8hJk24IRwO20rSI5GIjQ8fjUaxtLSEbDaLRCKBubk51Ot1m5ITJiHzhIt0KBSSsD0QCMhcYgKJY6xxSeZQyBuuVCpIJBJIJpNYXl6+r8/6YRNd0Vuv11Gr1YQMcPXqVYEgmH9iopfPUusZOyMSGltZWcGZM2fw1FNPScdMv9+P/f198dY1j5s00kqlIhS+fr+PZ599FvPz84hEIigUCsJA293dlYWETgW9bl1Kzh/Kceetjs1Y61JSnYShdxuNRm3wA7FuwiDhcFiSR85Epe6HGwgEMDc3J54NFYxGlkabXhCz9lQ6rqb09hjKsue2k2etJ5HO+h7V42JaRdOXuDh6PB6EQiFp0kVFYrMe3SccOMSDdSUqz8Xz6YWXRpiRE5UsGo0imUxiOByKJ0ZIRjN4uMhzQdfhrsYvyVrhNamsjNCCwaAsFNMszgKRUqkkbCvSc2moy+WybCRArrsTB+73+xL1ZLNZLC8vY3Fx0cbYyGQyuHr1qtgAPYb0qjudjhTHnTt3DqdOnUIwGIQxBuVyGY1GQzYUoXd/VCXs7eQ4df9YeNYaEnBiO7r/LLOo/GHXularhWQyKSuafgAcOACCZbIqkULDzwGjAnKvNV1BxYVCJ4Q4QLp3rfastFetJxW/98yzHgtplzTa9Dhv3rwp/TfoHbOYgeOmOdWEu4hbEs92JvzYj5zzhpsdJxIJdLtdMdS8J73wEp8klVC/x3vjHCHmrkN9jjm7PU57MydtJBlZ7e3tCfTFZ9bpdMSzJgNIG0bqMKEmbn47Pz+PZDIpn+10OkLl43WdzhS9806ng0gkgtOnTwsRoVqtYmNjA4PBQHqN0EbpRflB4dXAMRlrekdHhY401uQ78hiWljNkmpubu2XXF717BJWXQuXmA9YEdT5wfp54IjFL3aGP59dGg16CTlLwvp1GfJZchI05w0UxEonA7/cLDbPb7SISiWA0GknjLb0ou1yHG0poSEsXosTjcfHeSLFKJBKCcTsTksQntUeuvS5izkx0cqzZm6Zer8PtHm98wZoB3pcx480RnnnmGVy4cOE+P/GHT3SVn9frxY0bN+D3+4VlQ292MBjg8uXLUmiijSvnEJP7dLhYuAYcYsgce3rBjNLIz6ddYBn5hQsXEI/HZaPtQqGAq1evYnt7G/V6Hfl8HqlUSiI0RuecE9rr1w7bcRryiRtrKiqAW74UB4dhJRWNWCKxJMuyZPdw8mV1+1F22COOpJsv8brayFI0n1dj3/TsdOJQn0tTdfi/5lbye+oS+2kWzTvnAko8kl34Wq2WwAYa/+N4aWPNsaWHw54eegwYmWnWB2GwSCQilW3AWLFZCk5vnobf4/HYYDY2ASK7gAs6AIHOSPWLRqN4//vfj4sXL049G0RHzS6XC7lcTl6Lx+NS4VgsFrG9vS2GWheV0NFidHTq1Cmsrq5KQRQAiaCCwSAWFhaQzWZRq9VQKBRkIeZivrS0JHOATBCv1yubbJNFtLOzI3TDeDwuDibZZIC9WvF+ybHNqKNYEVzxdCc+CrnOLpcLkUgE6XRaQhfnaqbDZN37GIANuuBn9TX03zq00eENw9ujEqTO+3YyQGaY9aHo/AG501QKKiU3P9ZKxYWWtD29oGu8mpgzjatWTjI3aLRZHMEx1SEu54/TWHMcCZHoxUD3uOA9PfbYY1haWpIGZQ8qXH4YhPpJByiZTOL06dN4/PHHAQD1eh2NRkPYX1rXtPNFgzk3N4elpSUsLy/bCpp0rog5inA4LOXjAGxN3OgcMPdAr5uOAzcmKZVK0iiMMFyz2ZTy+Achx9LPml6m01jTs9YDow1ls9mE1+uVjD7PQY+VBtupBPTe+J4uyqFC3S5rS4XXnHAqH3uWHEWA14aei4822NMu+jkFAgEUi0XUajVbJ0TijIQ6SNGkkvM8TsPL0Fj3EtGRD8eGiUnCXYyeSMUkHKK52jQQnHc6f9Hr9WwwG4WMhW/7tm/D3NycRGrT7F3rhXAwGODpp5/GRz/6UTz//PO4fv06/uRP/kQaKTntgSYEsPf00tISVldXcfbsWdl8gHND5zsikYh03NRFcyx80iwjRkkcJ0bso9EIhUIBi4uLGI1GQsVkdeqDykkdS28QitNo6ebgbPBC7NqyxgT65eVlnD59GrVazWYoiXNSIRke69WVnpk2FM4EJQeXRpr8WkIwDJe5OvO8ZALwf+CQbkgDoMvRp130M2AlKPs7kF7lbMxFaEt72VR4flZz9/Xu6M6dSDSGSM+dCSbN5iBEohd0XUSzv78vBtrlcqFQKMi9MBkejUaxtraG7/zO75ROkdM+B8jeoT7Ozc1hcXERi4uLACCL95e+9CUpjNOQEmm7jLxWVlbw7LPP4uzZszLmXEABCIuLhXR+v1+4+tqhYiSWz+dRLpelQpmGmxg7mztZliX9Yowxwlrh+DJy4P/OyH6ScmzUPXqr/CJUomq1ilqtJpVI3Cyg3W4jk8kgk8kgGo2i1+tJDwndlpLJHhprejEap3ZCHPytE5VcOLglEAeLisykk7NJkBM6ceLWwPEmGU6KaMOnccdisSje1mg0QjwelwSghiLY9ZCl/wx7WWEaDAZRLpdRKpVQq9Uk8TQcDtFsNjEajSTJRfyb48lk4mAwkF4wVGonPZMJ6F6vh1qtJvkWTTU8f/483v/+99tYJtMuhBzYeIvPnX3gFxYWsLi4iGQyiXA4LCye0WgkbDE6SWfPnsVjjz2Gxx57TJhf+hkTwtDEAu0AAIf7beqCmVKpJAZZd2ocjUaSUK7X60I1JsSiaYFHjbXT9kxKjm2nmNsl2iqViuxCrMOfUqmEs2fPSlKJ3g2NMRMKLpdLsvEMlegZ0eNxYtNa+Uj9444SNNRcBHS1EztxvV3S0AmPzBT1ULS34fGMNzHlXnc0mFRMhqb0rJ3JYeCwRSmhCpYO6x1+eD2OMbFmjiOvq8dKN+Vy9vRgxMQOfpyz9PbT6TTOnz+Pp5566paiiWkWTZ0kJZc6z40m0uk0MpkMOp0O0um0dNckbkzn6Ny5czh//jwWFhbEaaI+A5DFXUdVnCeMvpwblnD+cOHlOQizccOBWq0miwF3teIiAtj7Hzll0nPg2DxrXWKub3p3dxfz8/OS/WXv6s3NTaysrMhxiUQClmWhWq2iVCohmUxKS9UbN25IYQWvx9WRTADnvXCQGApTyVlS2uv1pLeAxtA1Xu7EvY/yrvneNIuOMNzu8a4ulmUhEongsccew+c//3k0Gg3hwu7v78PlGm9WTPpmq9WSbbLoFRGeYuKSCSUaWe7BSY+u0WhICM5Oar1ez8YkSCQSNhiOm+dyN216z5xXTJLG43E0Gg2cPn0a73//+/HEE0/Yoq/7WYb8MIoxRgqSYrEYms2mOFU05MlkEqurq9jY2JDIKxaLoVqtigF97LHH8KEPfQjz8/PSHkKTFAhP0cHT/WXIyybcxX0W+/0+ksmktE5mpK43S67VasIqASDtKlKplEBqvL4Tij0uORbMWnssgL0+f29vT7bDSSQS0imr2Wzic5/7nFC6uANIo9FAuVyGz+eTsmFyMrVhBQ4bATmTkMSpqNTsrEeGAld9bbg7nY6thB24ddcYnWjQq/q0c631IspnRa+F23Gx+fzm5qYYw83NTQSDQalW293dxdLSkjBziFHSUy8UCigWi2i1Wjhz5ow0b9rZ2cGpU6fEkG9ubgKAbEoxGAywsrIiUMuNGzek8jESiSCXy4l3zwQUF3rLslCv19HpdFCr1fAjP/IjWFlZkfcIz027EC+mEd7c3BRYgdTZUCiEU6dOCdQQj8clzzQ3N4fl5WW88MILsmEEE746KiJMysiaC7XP57M1YiIuDgDRaBQ/8iM/gscffxzxeFyiL/LAySzjPCQziTx9wjVHJRp1JD9pORbPWoeCDA1588yo7u7uSqaeuCIbrJTLZfGgWP5JSp/X6xUvi2Et/+dAMttPo8GQWCcSGNbyNYa4uspJt1Sk8HvogeJrM2N9tOjx7/f7iEaj4qWyj4tu2gNAIhvODYpuW8n36PVQgXguJwWT85GLMAteeF+km2noiwlnzhNGgsYYLC8vS2N9Hj/tURWFiV0m/fb396XATXumhDr5XrfblU1Ezp8/b0sSaqwYOKT4MeqNxWLIZDIoFovY29uT3Bh/2D+Gjb6Ii2sKMO+Z84DjrTdpJqzzdkb5OGCwY6lgpBwFg9C72dnZwdzcnDxsUrmuXr2Ka9euYWFhAcFgEJ1ORwju7MYWDocxPz8v2BKxMDI/nHQ9fkbvHELl0+9p743bfenv46TuORMJzkTjtIuTB8sfloOTMkWlo1JxsdPFUxSGulRuKpBOCDPfwCo34tT0yOldczGmQTfGCOVLt9NkUY+eO/QK2eOGxRL8rtOOWQOH7Cwm5nTZNhdmzfii7q2uruLUqVPSWI27yGssm+PGY2mss9ksisUi4vG4bGyhWV+ZTAanTp2yVarSaBPqInTCZHW73RYDrpu7OZ2542aCHaux1okinRio1Wq4evUq1tbWRKEJc5RKJRSLReRyOXkofr8f586dEw4lqxudXhYwNszs6saBpKLxR3vZxK2p6IRHjlI+J1NA45PaUM88azvPmvSt4XCIZDKJ3d1dDAYDBINBUapms4n5+XmBGarVqtC8uGEFdxPXRSscK+7JR1YIx6BareLy5cuyaPT7femZzMWgWCwilUrZ+NTsD9LpdCTK6/V6Ypx1syf9PZ1J0WkVtmwgK4QUvNFoJBsSFItF9Pt91Go1MZDRaBQXL17EmTNnEI/HsbOzYztWY8Ok79HrTqfTWF5eRq1Ww9bWFvb29mSBpdccjUaxuroqG1wTPmHbgFgshlQqJZvzckf1UChkY4VpfXdSiLWtm6QcG8+aKxVf05l29hne29uTXg6aBL+4uIhSqYRmsymNWy5cuCC7YLOYgQ9EJxSdDezpadOw60SjhkVowBn2MHTWRlkPgFbSo5KN0yzOyEOXm/t8PiwuLgqVK5PJSNKG3gyTPGwoz+w9FUOzfhqNBnZ2dlCv15FMJgVXPHXqFPL5PG7cuIFXX30Vc3NzYhBisZhQNtPptDR5crvdkvjkvAEg964rLIPBIJaXl2XucTHifWnDMo2iDWksFhNjR2iiUqmgXq8LlS+RSGBxcRHZbBbnz59HNBoVj5jjo3n1WhfpEHLecZEIh8NoNBoCaWYyGbEfhUJBovvHH39ckpDcEo4OGHdTr1ar4hhoO8Ix5lznvD8RmLX2NokrM3ygsSbHmo3lKboRE7FMZo11Fz7delEXwnC1c7IydBGMZobQaGvjTeyan7td1aQTHuHv40ounGRxdqfTbVD1cwYOC42ILWvsGbDjlLqEmMa02+0in8/DsizkcjlsbGzIPNO7wxD66Pf7ti28uFmu7szIaIxziPN7YWFBviMpX5QZDHKYX+Ai3Gg0ZM9Fji0putxcYHV1FbFYDC6XC+12G41GA8Ct+7lS9Gs07JwLmuBA4gIwXnxpfLnhCBs2seHYUf24CZkedQ/aOTwu/T+2eljtKXGSU7hiVatVZDIZALB5JWxpSgVLJpPSqY8rGfEvrqSAfSshpwGl0aWHrxtK6Xvka9rwM4HxdpNlZqgPxWmodJtJvfsGE8jNZlPocrpysd/vSx8Izg1i1WxzOT8/L5n/Xq+HcrmM7e1tbG5uYmdnB8ViUa7FqI0eMTcM4MYDHH8Ath1HWFxD7i3n4OrqqsxbJqBu9wymTahnfIbcdi8ajcp+icw9sNoxGo3i1KlTMvbc/guwNwdzRrH8XavVZDci6jT1ttVqCeZdq9WkzD2RSAgkFwqFEI1GhbbJ6IAb+xIK42Kt8zFH5aoeep41kzi6kQsw/kLNZtP22Vwuh3Q6LaEHj2OyAADC4TCSyaStpSoNqH5NNwjXVY06HKVB1kUTztao2rNmZy5tqG/nYev3ZwbbLqRjERvU1MeVlRVcv34d7XZbFISKqHdB189VQ2xzc3OYn58Xqme1WkUkEsFXv/pV+P1+LCwsYG1tDaurq5ibm0MqlZLSZhp93QNGY9vdbld2stEc3kgkglgshqWlJVtp+rQbaC37+/toNpsol8vY3d1FuVyG2+1GIpHA2bNnUa1WhRHEbd96vR7C4bDsy1gsFnHjxg2ZD4888ojNOSObh5sY7OzsyPZg+/v7ArkAh4Uz/X4f5XIZ+/v7UlAHwNaKVzeK4ga93KiZXr9TNAOOP5OGQ4+10wzDXCYHtMJZloVyuYxyuSzcRZaZc/frRCIBALJq6j4RAAS/1Kuss/KNg8uFQCsdjTWNBwdOs0icOLTGYp2DMTPUby80htozWVpaQrFYRLfbtZWbc4ypbEclcDUH3rIs2TtR87JZ6s7qSC4I5Gtz4dY4OLFxzl/OA4/Hg3a7jbW1NWEU6ATTTA5FszCq1SoAyN8+nw+pVAoejwe1Wk0KjJhwJG2THi3hLeq6rlwslUooFArI5/PI5/PY39/H/v4+isWi9PCwrHHBHGnDhGEYURGL1j/Aoa5rbF1H4xqzpuj5eKKMtf4i9JB0mSZ3Ka5Wq7L1UqvVQqVSEawSgHi5ZIcwaxuJRGzKrVc0jSFpRXRi1hrrJh7J1zSXkkrpbPijEwszubPoBIwxRirc6vW64Jv6MxqiIvOHxt45NkxSc64R66ZHpcedeYrBYIB2u20z5HrB1fkPzhMWbLwd8+OoPMc0id4dnEaNvVjYq4ULZTQalf0Ta7WawCTValX27KRuEucGxot1oVBALpfD3t4eisUiCoUCyuWyFC7pHujdblf2fSWGrQvgGo2GJCS1oWVPEb3Aa9jDOWdOFGbNicrJTGOpG4vTcBYKBQlX2TIxl8thMBjIBpfGGOnQx6zt0tKS/CwuLkozehLZLcsSipdWWnrXRyUYmZgg7q277Gmv3Rni0HhoT28mdxYWtKRSKTQaDayvr0uva8uybBAEjTKNMRv06GIGGsdgMCjh6mAwkF4kzhJly7LE2+JOMyzUckZe9MCi0aj0qng7mXZjzSTvaDSS7nk0lppZQyou2SGXL1/GxYsXZcfzSqWCdDotOlosFmUu9Pt9vPbaayiVSqhWq9je3hbKHjs76vFm1Sl1nUlswrOEUQiNUed1wy/aAicselRU/dBj1vSGdDWY5jJqIrllWdKLYX5+XsD8RqMhrStJnNdVioVCAZcvX0YoFEI8Hsf58+fxyCOPYHV1FadPn7bxcZ09j7XnrL027bl1u13psKY7eemHrwdrmila70SodPPz8xiNRlhfX0er1RKjSWXTeQUyhuLxuCyQOpTlIszdaKrVKra2tmxFT0weslrV5/Mhm80K48jlOtzRnHOl1WqhXC7jySefxNLSkvSuAY5Wymk21MC4rcTi4iLi8Tjy+TwAyCYDTPYSjvJ4xjv1tFotXLt2DdlsVuhyhUIB2WwWo9EIuVwOf/RHf4RsNotgMIh6vY6vfvWrKJVKqNfrYuBrtZq0NNWFTxx3Ukl1rQU3Quh0OuL1c875/X40Gg2J6jWMpwvjnAv0Q2+sKXxAwCHd6nZhZqfTwe7uriQViVcS8Kc3HY/H5ffNmzdRqVRQqVSws7MDv98vEEs8Hkez2ZTwhVlcrbA01NrDprB1qhZnIlFziWdY9b2Lfmb0hJPJJEqlkhSYaH41W5TWajUx2oAdnuDYWpaFnZ0dNJtNNJtNyeRzAXe2ysxkMuKBkcFE5dPtB1qtltQE6GiLopV12o11Op2WyIjGkAujhqiA8RZbHDu2vGUzL9L3SqUSNjc38corr0gBSyQSwfXr12W3dPYYIlzh1E1n/YTG1elhsx8Rj2elYrvdFhiW4szBHbfc160sOMm1GGOkGfjp06dhjBHDq8ntS0tLmJ+fRzqdxuLiIvx+P9bX15HL5TAajftkczE4deqUVCUyYcHsvhN30lVJfPismHMWvGisWntVM2P9jQmfm8/nQzgcRiaTwdbWFgBIn3NN22u1Wtjb20O5XBZ6lub0awPJaEwXtOjqRFL0dKkxIzd6Y1TkdrstxoCfP8qrnnboQ8v8/Lx0smw2mwI/6F4tfF7shEejyD0vCUdUq1Xk83mEw2FcuXIFjUYDfr8fFy5cQC6XE09as7m0zmq95nWAQ3ohr6kNvmaRcQ4QfnOO8+10f9I24diNtfZA9O4uerL3+33s7u4il8shk8kgmUxif39fWiq6XC4kEgkYY1Aul7Gzs4NCoYB0Oo1HHnkEZ8+etTE8dnd3bYOijbWuUCRdiIrr8XgkyVCv1wHcWsHo/G7OMGgm9yaEqNhpLZfLSahJKITvx2IxuN1uZLNZMQD0wknlI0zC93RjHh326kZd3DeRMF273ZbEd7FYRKPRkJJ4GgLgVmWcGepD8Xq9KJVKQqFrNBrSu/z69eu2DY2JBdOzJrmA3faq1Sp2d3cFNtnb28NoNEIwGEQikUC325VdnQhfcUHlmDCX1W63pb95vV5HuVxGoVCQ7nrhcFionawJYGRHJ85ZLOeMpo7LFhxbBSNFPyxyqXXrQj3x2QeAW/rwwYdCIbz22muyApPyRViEdf46NNZJTBpqrtSaX63hj+FwiGKxKJt46u/gXGCcYfjMUN+baM40nyd3c2m32wKjsUSc/Nd4PI7BYIBSqSRQBsuV2WYzmUxKgoneMUPcXq8nRThcEKLRqI1pUKvVUKlUUC6XBcdkeOysHXBGWDMZCw00GyHpIplQKCTGkX9Tr1OpFPb39yWy6fV6CAaDUhbeaDSEb722toZisSidOEn9A+xMNK2b9KRdLpfg3GSIOPvOAIfUX53juht9P45o+1h6gziNGw2nLnqh6C/EAR4Mxn1wm80mYrEYTp06JZScWq2GcDgs2XudxNT3oJOI2lBrzJqGmgPFhkD0oJyLjjOpqCGSmdybaPoef8jqqNVqKBaL0mqA3fM084Nl4xwnXYFKmIPjzRwEoyrCXoTZdJEFQ3eWGBNOoyeuWSf6O8wgELvoLfFIhaNnSuiLezS63W7pBcKiFZfLJUwgtj7NZDJ48sknUa/XYVnjvRHdbrdwrSuVig261H8Dh/1K2GJAd9zkXKIcZcOOMtSce0fZuxORYNRerX5guq/HUSsPQ896vS47C2ezWTzxxBMSqoxGI2QyGWkWD0AUVochNNRsd8qOe7q8nPxcKnq9Xkc+n5cEg95814lb8zgtM+z63sS5GHK3jkqlgmKxKN4vn3swGJRjAoEAEomE9Jio1+tot9soFou23ticg3QCNDOIO8/wfzJPWLZMSEx7XLpQhslPPS/ejiEyTcKybGOM6C2bIrFVKZ0uVgayI2I+n5cGWeS1LywsIB6PI5PJCKuEEEkmkxEmCMfKmfwlnMqkcTgcFnvAPkRMaDMB6tRnHnuUjjuhUtqWScqxtkg9aoUhDsnBdH5xsjsWFxcRCASQy+Xw27/927Kf24ULF+RvbTC1suhkIpMVTo+aBp79IUqlEnZ3d5HP54X2o6svmajUHF39/TRzQa/Q0yp3Mlp6ASRTiDtT7+/vw7Is6euRyWSwvLwsRQwAhEbFawyHQ8GVyeXlmLEHOhdxALJBq9/vl7CYxVdkGZFGyvv1er34zGc+A2DsWLznPe+ROUycVEeW0yyk3tKzdrlcsgju7+9jZWVFksgf/OAHsbm5iZs3bwrcxERuIpHAo48+irW1NfHE6YUXCgVUq1UkEgk8/fTT6Ha7uHTpEur1ujhjeid0zUbTUVitVsPKyoq8zh4hTCxaliV2gws1gCPtz3HKxI2108NwihPzo2hlYXKRrTMZpkSjUWmVyO5YZAHwXLqDHhNJzsIKhj/dbhd7e3sS9pbLZRu/lsbYudsNYRC9yurqzJncmmzRYaUzKTMcDlGtVvHqq69Kpj8ej0slGnDYWkA389I9Xwh9uFwuoYDqSlbN+IlEIrY9FilsLkavmtAHxe12Y3t7G5/85Cfx2c9+Ft/6rd+KF154AdlsVnBvPben2WDrvQk5ZpZlyQYjjz76qPQG1w5VOp2WnJHL5cLp06elfxDhMG1saYA9Ho946s6mT1yw2QaXXjbx61arJVEbIygeC9jzEUcVvh01zscx9se2rRcnrROf1saaoj/DkLbZbGJpaUk88XA4jHg8Lr2OnRsMaOhDE9oZDvMBEzsjXadYLNoSSNooO42z9pr0d6BB5+dnDejHovFCZ8RFb4UbIm9vb+OLX/yi9AgBYNtnE4AklAlP0RCwIQ+7t1GBdc8X8u2NMdKoB4DtM2x1QKxaJ5sI3XU6HWxsbODKlSuo1+twuVw4d+6cFICw+RTn5rQKdYLsHG3wyuWyjcvOxO9oNBJGTrPZhNvtxtLSEqLRqHjVzDkwUQwcNmkKBoO2507b4HK5pFc1YOdYkx3EPi/OeaG/D3DnjQVul4+bhBxrUYzTg6aRO8qQa3y73+8jl8vJJqgMS7hRAfsRa8PMh693Lm+1Wmi1Wjb8iuXHlUoFhUJBMEmusEclD49aSTV2rQn+LKufdtFJPOBwAwrCSkzm/umf/im+9KUv4dq1aygWi3j++efF2KZSKezu7kqBRCgUwsrKipSjE3MOh8PS9Ite1nA4FG+b1yX85WzqRWpeqVTCzs4OSqXSLdx7SiwWE+bBpz71KfzFX/wFzp07h8ceewwvvPAC3v3ud2NxcVFa/06rsJw8HA4jEonI/wAksQtA2FeMcPjsCZ1whyhiyfSoW60Wdnd3pVUuW6xqlhbHzuPxIJvNIp1OCzRaKpXkXFzouRjTkXDqPRd1/d79XJCP1aocBfLTwAK37gSsDerGxgaefvppJBIJgTtoFPWAaAya0AcTQ2wkzwKJZrOJXC6HUqkkBHgqLP/XHjPvVyccGBnoCineC1f2mWd92H+YoSo90263i89+9rN49dVX8dJLL2F9fV04squrq0gmk6I0brcbmUxGjOmNGzcwGo0wPz+PbDZr897C4bBEW2zUw/JztjogG0EnmtmOk4ktKjGdDaeDMRwObbvY9Ho9XL58GW+99RZ+//d/H0tLS1hdXcX58+fxT//pP32AI/BgRVMvCWMwsc/umqT2cfMHUi0JUQHj4hq93yGjcxbXcPd0vYktF+x+vy+tCU6fPo14PI5CoSDJaNoNUn+Bw3l7VO7tdkZcy3GSDI6dZ+18TRerOL+UDpmZiCBWRfqUDnH0yqir06iQhEGazaZQscrlsm3/PQBSGHOUN83f9Nb09fm3hkdmxvrW0nxg7DHlcjl85Stfwde+9jVsbm5id3dXGvsHg0FkMhkxjkwwRSIRAJAFlbQuGmcaVGLfGqaih86cA3CIpeqeIqVSSZKJTDwfBdXRs/P5fLIwA4eFPe12W7YY293dPfbn/DALFziv14t4PC59qpnnYa8fOkL8bCQSkWQx6yCcuS1G0sxbhUIhWJYl/Xw0xY7wSCqVEocPOIyAOX8I1TBi14lD/feDZHsde7yuvxxZFc5yU/1ZKhvDHMIfPIarIT+vPSTS9Ojx0Mtm1piv8170PZFMr0UnxJzQBiejTlBRoafdWOuIiZViV69exSuvvILf+I3fwM7Ojngz/AmHw0in0/K8WeGmcw39fl8y/alUCrFYzIYlMsrS/T1omJ39YDhXuINJrVaTSEvvPATglgWBGCkXAb1As2CnUCg8gCf/8Aj1gAVL8Xhc8gIaTtBODj9rjJHcAQvhtPfNnBZxbMsab7JcLBZvaWusWyqTHWSMkS6Auosj54smGQDHx5u+Vzl2Y62xa53wozhXKm0EX3/9ddtGBMzih8NhW3HCaDSSggc2OCcWxgSiThjoTDWVV/Nm+TnNZCDWyvd0KbTG21mwMe3CSb+9vY1PfOITePHFF3H16lUxzKR2EYsOhUJIJpOyRZbGN2kI2cypUqng5ZdfRq/Xk01yuQ8jx4tziEqnWwyQ+UMF507WGmPXRoS/mfjkeaLRqK1SEjgsd6dxmVah5xoKhTA3N4fTp0/bKhKpw/V6XRyoYDAoz45RzpUrVxCPxwGM58JXv/pVbG1todPpIJlMwrIsbG5u4urVq3jjjTdk8QTG40TPOxQKCRWUVavtdhvBYBDpdBqxWEzmCBPbTl4+f+ucllOO0/M+tqIYwN4+dDgcSl29U7Th5A+9642NDTSbTSwsLIhSkmNNj0cnk7i9E2l7hDeocEeJJtJrA61ZABSNm9OwE/c2ZlzYkUqljuGpnhxpt9t46aWX8PLLL+OP//iPsbW1hUajIUwNzfBwuVzw+/0IBoMCLTBiosL7/X4pYvB4PGi1WiiVSrh+/bp4bfF4/G0ZGKTisbsaPTddtq5ZPTTWzqpbGnTNOtEJSy4Szq6N0yZ8Fl6vVxLDzWZTIC9GVKxEdrlcMv5MFIbDYVy7dk3Ky+v1Ol577TWUy2UptOEWXSw1p/fMRZ4c/Ugkgnw+b2tzYVnjnumZTMbmuTPX4Yz8dWsEZ7sMiv5/0p74faMtEKe6HabtXKmIFbXbbeTzeRlk0rP4GeCQjkdPyukhO6k8/P+o/h46sURjzZWU1CInJMLzER8jM2Fa5c///M/xla98BZcuXcKbb75p8zqBWxvd0Gul56rHS0MSXKQB2KoL2+02ms2mrfOeNrij0UjKzfmbSSbuv8jracxbJxc1p1vDM/rz+poPEtt8GEQ/v+FwvBkyWxcTuiAcQWeHCzOLYVqtFm7evIlcLidJ5q2tLVtf7N3dXVvxEoV6nE6nMTc3Z9s7kdG02+0WfdVRtiZAcFy1c6jHWNuTo1hjk5RjKYo5qqqHxloPoJ7UztJtPiAqWqvVEoVmdpgPi4PNTPHKyoqt9SXfoxfs3F3EyUoBDot0NC9cM0N4v7pAgxMynU5P+rGeKPn4xz+O9fV11Go12ZWe4pzcLBvmuHIMNSWS/9MDtiwLyWQShUJBKtDYZY/v63CYPGv+6J1AmIikJ8X/gUN4S/+vPW1+ntcE7PtMTrNwzIhPu1wuRKNRzM3NSa9yAAiFQpJs5ILc7XYFKul0Orh27Rr8fj+63S42NzcBHNL/dFsBziOt59lsFqurq0gkErbIml41c2LA4b6RPJcuhqOx1puR3G8x0+4BzGQmM5nJSZDZflQzmclMZnICZGasZzKTmczkBMjMWM9kJjOZyQmQmbGeyUxmMpMTIDNjPZOZzGQmJ0BmxnomM5nJTE6AzIz1TGYyk5mcAJkZ65nMZCYzOQEyM9YzmclMZnICZGasZzKTmczkBMjMWM9kJjOZyQmQB2qsjTG/Zoz5xQd5DzN5sDKbAzOZzYG7kzsaa2PMDWNM2xjTMMbsGWN+1Rhz37vrG2O+1RjzqjGmYowpGmP+kzFm+S6Oe/7g3hvGmKYxxlL/N4wxp+7H/R/cy88YY64bY2rGmB1jzD8xxjz0u+s+LHPAcU+/ejCW5+/isw/THPhZY8xrxpi6MWbdGPOz9+va70QepjlgjMkaY/7DgS0oG2P+/V0c89DMgYP7ebcx5s/V8/zv7nTM3XrW32tZVgTAuwG8D8DPH3Hx4zY6lwB8p2VZCQBLAK4A+OU7HWRZ1mcty4oc3P8TBy8n+JplWRv87H34Dp8C8G7LsmIAngTwDICfPuZrTkoehjnA63wLgHN3+/mHbA4YAD8CIAnguwD8pDHmB4/5mpOSh2UO/DaAHIDTAOYA/L/udMDDNAeMMRkA/wXAxwGkAZwH8Ok7HXdPMIhlWdsA/jPGhgYHq9PfNsZcwdh4whjzPcaYrx+sep83xjytbvJZY8zXDryKTwII3MO19yzL2lEvDTH+kt+wGGN+wRjzm8aYf2eMqQH4MWdIZox5wRizpf5fMsb8ljFm/8Azumtja1nWNcuyKjwVgNE7/Q73Wx7kHDg43gPgnwH4yUl8nwcwB37JsqyvWZY1sCzrMoDfBfChSXyX+yUPcg4YY/5PAFYB/KxlWVXLsvqWZb30Tr7P/Z4DAP5vAP5/lmX9e8uyupZl1S3LeuNOB92TsTbGrAL4bgD64Xw/gOcAXDTGvBvAvwbwNzFeMT4O4PeMMX5jjA/A7wD4dQApAL8B4K85zl858Jpud/1TxpgKgDaAvwPgl+7l/m8j3wfgNwEkALxtOGWMcWHsHb8MYBnAtwP4GWPMdx68/y0H9/d25/ihgwlRwNiz/vg7vP/7Kg96DgD47wH8uWVZr0zkC43lvs4BdS4D4HkAr3+jN/4g5AHPgQ8AuAzg35gxHPoVY8xHJvC17ucc+ACA0sEiljfGfMrcDQyjt9Q66gfADQANABUANwH8SwDBg/csAN+mPvvLAP6h4/jLAD4C4MMAdnCw4cHBe58H8It3uocj7ikF4O8C+MA9Hrd2cM+eg/9/AWPF15/5NX1PAF4AsHXw93MANhyf/zkAv/oNfIdHAPxDAAv3euz9/nlY5gDGHtVVAHF17fMneA78A4wV3v+gx/gEzYH/z8H1/q8AvAB+8OCeMidlDgB46+Ce34dxVPFPAfzFnY67W2zm+y3L+qPbvLep/j4N4EeNMT+lXvNhjDFbALatg7s9kJt3eX2bWJZVMsb8GwAvG2OWLcsafCPnOZDNO39E5DSAJceq6Qbw2Xu9qGVZV4wxr2M86T92r8c/AHkY5sD/BuB/tCyreg/H3I3c9zlgjPlJjLHr5y3L6t7LsQ9QHoY50AZww7Ks/+Pg//+vMeb/gTGU9Lv3cB6n3M850AbwnyzL+goAGGP+AYCCMSb+dnN7EtQ9/dA3Afwjy7IS6idkWdZ/BLALYPkg9KO8kwysB+PkQuwdnAOw3z8ANAGE1P8L6u9NAOuO7xe1LOu7v8Fre3APibKHWO7XHPh2AP+rMSZnjMkdvPYFY8wPvbPbv79zwBjzNwD8PQDfblnW1p0+f0Lkfs2BV3DreE1C7ucccH4H/v22GztOmmf9CQB/yxjznBlL2BjzV4wxUQBfADAA8NPGGI8x5mMA3n+3JzbGfMwY86gxxmWMyQL4fwN4ybKs0sH7v2CM+dMJfIevA/huY0zKGLMA4GfUe18GUDPG/F1jTNAY4zbGPGmMed9dfocfN8bMHfx9EePQ6b9O4J4fJjm2OQDgAsY4/7sOfgDgewH8J+DEzIEfBvCPAXzUsqzrE7jXh1GOcw78JwBJY8yPHjz7H8AYN/4L4GTMAQC/CuCvGmPeZYzxAvh/AvicdUg+OFImaqwty/oqgJ8A8M8BlDHGF3/s4L0exuH+jx2899cxpuCImDHn8PnbnH4ZY7pLHcCrGDMp/qp6fxUHA/YO5dcxxhFvYEyn+STfsCxriLFxeBeAdYyThL8CIH5w/88bYxpvc+4PAXjVGNME8IcHP39/Avf80MhxzgHLsvKWZeX4c/BywbKs9sHfJ2EO/CLGSbevmEOO77+awD0/NHLMc6AE4P+MMcGginGE8n2WZRUOPvLQzwHLsv4YY73/AwB5jBlhd4wOv2l2NzfGfB3jsLL4oO9lJg9GZnNgJt/Mc+CbxljPZCYzmck3s8waOc1kJjOZyQmQmbGeyUxmMpMTIDNjPZOZzGQmJ0Am3rDkox/9qNXpdDAYjOtUnn32WYRCIXg8HgyHQ/R6PbjdboRCIQQCAViWhdFohOFwiNFohNFodMs5XS4XBoMBXC4X3G43XC4XLMuC2+2G2+2Gz+dDIpFAKBRCMBhENBpFPB63XWM4HKLf76PRaKDT6aDf72M4HKLT6cAYA2MMXC4XvF4v/H4/vF4vhsMh3G43vF4vAoEA5ubm4PF44HK55HyDwQCDwUCuwWqjixcvvi1n8ptZ4vG4BUDG+cMf/jCy2SwSiQSWlpZgWRY6nQ4ajQb8fj98Ph/cbjcAYDQascoLfr8fwWAQbrcblmWh1+vZnvloNJI5EQgE5DycEzxfp9NBsVhEu91Gr9dDMBhEKBRCNBpFIpFAIpFALBaD1+uFy+VCIpGAyzX2YzhPRqMRBoMBPv3pT+OTn/wkXnnlFQyHQ7m+MUbmPKVWq03tHNjY2LB2d3ext7eHjY0NvP7662g0Gmi329jf30e1WkWn00G32x1X53k8ooNaH2OxGFZWVhCNRhGJRBCPx2GMkXH1+XwolUooFotwuVx48skn4ff70Wg0MBqN4PV6xWbQBg0GA3g8HrRaLQyHQwDjeVKtVlGv11Gr1dBqtdBut9HtdjEYDFAul9HtdjEajXDq1CkYY2SeNptN9Ho9DIdDsR+0B2+++ebE5sDEjXUgMO7Jwonr8XgwGo3QbrfR6XRkYADIBB8MBvLQOFgUDlyv15MBpELS0ANAr9eDz+ezKTWVzLIsDAYD9Pt99Ho99Ho9UXYaWp7XsiwZCBprnqNUKiEQCMDtdh9pOHq93qQf54kUKhMAWejC4TCi0Sg8Ho88/263C5/PJ+PKY6hcXDS5WAOQxZLGnYspP+s01lxAfT6f/E2nod/vA4DMCxpoGgFtQPi9Ll68iNXVVdy4cQP5fF7mHACZO/zsNItlWWi322g2m+j3+/D5fDLGTocsGAzKuOr3XC4XgsEgvF6vjIXf7wcAcZbcbjcikQgsy0IulxMDHYvFxCZ4PB6ZE3SmhsMhPB6PzfZQz6nHdN663S6azabYC+BwHvZ6PTkPfyjOxfudysSNdTablQHhzdObBcYPwOVyySACsHmlfJ9GcjgcigdDBfB4PIhEIvJZKhOVhoo8GAzEyI9GI9vKzff1/zyexpd/9/t99Pt9dLtdWWi63a4Ya+1lG2PkM9MqjDyMMYjFYpibm0MqlUI4HIbL5UKj0RCvJhgMwu/3wxiDbrcLl8sFn89n86qdSsIF3OPxwOv1ijJzLPWCzrHxer0yrq1WSxQ9Ho/D7/ejWCzC5/NhYWEBrVZLIjfAPq9WVlbw5JNPIp/PY2dnB6FQSOap3++XxX/aJRqNIpPJoN1u4+rVq+JFdzod20LOxdzj8cCyLNExj8cDv9+PcDgs5/R6vQiFQmITOD6RSASpVAq5XA6NRgPhcBipVArValXOxcWCURAXkF6vh1arJYaXxpdOmd/vh8vlQigUksXY7XYjGAwCAFqtllxDO463QwneiUzcqly4cAHA2MjV63Vsbm7KAHm9XgCQh1atVmVia4hDG0+GuvwMFZUDzcHmuXu9Hvx+v3hKVF4uCADEg7YsC/1+XzxpGlsaGq70HCCGV/TaqZg8Nz836RX1JAo9Xp/Ph3K5LMpIpaX3k0wm5TnTA2LIymfrXAA5rtprokF3u91yLJVGR0ici6FQCMYY1Ot1UchOp4MbN25geXkZHo8H7XZbvHV6XcYYPPnkk2g0Gvjyl78sc4oGADhsjjbNsre3h3q9LkbxqaeeQqPRwN7eHj7/+c8DgBhFji3nACMxwpipVAqxWAypVMoGmWkdtywL73nPe7C9vY3hcIhAIIBYLCb2w+/3w+12o1wuo1KpyNzkuHLseH3aA0ImtBP1el3ugfPKsiwEAgFxOhqNBrrd7sQj7Ykb67m5OQCHnifxYRpMPhyuZBQaOiduRa94NBqJtzQajWyKwcGidwtAQl2/3y/KOxwOBYOiElOxeB7txeuVkX93u120220xzlxYaJz0a9MqOp/A6CUQCCASiWAwGAhskUqlbJ6pZVkCewCwPX+9YOsfvdDS49JRFOeONqAMb2kkWq2WeFL9fh/V6riXDr0lfaxlWUilUlhZWUE2m0U+n5f5zLkzE6BWq6HZbIr+l8tlFItF5HI5cYpo6KjLxhiEQiGEQiGBtZhXCIfDEj1xjvB44HBcer0eut0uNjc3EQ6H4fP5ZKz7/T5arRY6nQ7S6bQtV0aHwOfzCZSrozIafRpoPV/5feggcAEiFDcpmbixTqVScLlcaLVaKBaLNhiDQgWiV0Lhl6Yi6tBWh5fEsHlewh1er1de50LBQQcOH/7tPB8dvtNY8755rwybOp2OKDjxMIbo0+5V0UOisfT7/YhEIohGowIxMFR1uVw2r4bPUCcatehFHDj0YukFaciL467PxWM5bsYYtNtt+fxwOESlUoHH4xEslQ4A52E0GsXc3Bzm5uawt7cn84TXnfbFGgDK5TLq9ToqlQparRYKhQJyuRzy+bzgxXyeOpEfDAYRDoflf+Y66LXSqPKHxhaAQBOlUgmbm5toNpsIhUICgZBMQH1lhEwIy+v1ipcMwAa1cEzpYPK6tFUc+8FgYPOyJykTN9Z8sATzdYLGaTQB2LwonURkKMLPc2AYjmgPmCuyzsS22215kAxzeR0aEueCQKXmqsnrcHEJhUJot9soFArY2NjAwsKCfNdkMimYKwd42oVeUCQSkdCRLB0yMgDYvF8udvxb5z+0aI+MY0altSzLhh+PRiMEAgH0ej202220Wi1JXPE+y+UyRqMRIpGIMBT6/T7OnTtnY5602234fD7E43GcOXMGr7/+uizk9NSnfbEGgNdeew35fB7lchm1Wg3Xr1+/ZdGjDAYD8ai1rSAU0u/3ZfHk+263G+FwWHSV0fb8/DwWFhZw4cIF7O3toVKpoN/vC5uL9qlcLqPdbtsSknTwmKfqdrvCNKIBdrlcqFQqErk3m00Eg0GBVDgnw+Ew4vH4RJ/pxI219nicWVW+R0PJL0yl0qERvR1+nufkbxpwekfAWFk6nY7tukxY8XOWZYnH5Pf7xehzddQRgN/vtzFGdLJyZ2cHnU5HJlmr1bKFXNMsXFiZQCStzhgj0Y72rADY6G8aznCel2Otx0XDHlx4dS5CR3GMxGq1mizcxEa73S729vYQCARkXhSLRYRCIYFIyEbw+XxYW1uTEJmsAWeyelrlz/7sz1Cv18X4EVpiTgE4zEnRUfP5fAiHw7KQGmOEFtdqtQQWASC6r+cEAJkLoVAIkUjERlzgPOA46twHGV4ABOYEYEtmBoNBJBIJdDodMfTNZlPmns/nQyQSQaPRQKVSQbPZnOgznbhV4SqoPVPAbqA5QE6F4/vA4UPX2KM21gT+dbhNQ9tut20ZfK7YfJ8Tg9AJFwkAtmtqxSPeTe+91+sJ95KJLGJV026sAdiet04c6nGlIuhchjbSej4Ab28AtVfN39qwE8ukc8CcCceUUZbH4xGqlsvlQrVaRSAQECPPBd7r9WJ1dRWhUMjmbMwgkLFsbW2Joda5C8CenwIgho41ElzMXS4Xut2ueLg6EiKtj5/l+3SotA7yOv1+H+12W+5JC+eh/jxwqPuMmDXNkHOm0ThssMeENOfWJGXiVoUZdCYEmVzUOLDmJmvRCT+tsPSQnA9RG2tdrKIfvMfjEX4vV2HSbZhM0A+e5yV+pfm6DJ95vmq1KsrPUE4nyKZVmI1naKjpdBSNM9Ob1mEkP68TQNpwO+lSDF31PQCQa3c6HTSbTcHMea12uy1MH4/Hg3Q6LRQwjmkqlRJ+OI2A3+/HmTNnEI/HbVRAfrdpN9qVSkUWMWe0SUdLG+xoNIp0Oo1sNotOpyM5BZ3MZ76Jx3COMXJnQhGws3P4PwCxR3xtMBig1WrJ5/Qc4zEau9Y5MJfLhUgkgu3tbeHqJxIJBINB8bonKRM31o1GQ1YWhod8IJpOBcCWjdU4sVZU7Ylrmpz2ZPREoGIDhyt4s9m0hU3a06Ox5spMKiBDYb3Ks/rS6/Xi1KlTeOmllwTbojEIBAKCxU6r6NwEAFuFoMb+yGt1jj1gH0f+r5WGiqALnyh6sac3RN625sJrKIZKTGyz1Wqh1Wpha2tLkqGRSATNZlMUNpFIIJPJCP6pC2SmXegtj0Yj8TA124vPye/3w+/3IxaLIRqNIhwOC+0ulUohHo/LwtpsNtFut+UapVJJHCQAwtHmvGH+isfToHY6HaEVcsHmeWlfer2esJo49+hsOgv/4vE42u02isUims0motEofD4fUqnURJ/pscAgXFGpEMSM+B4ViSsl/9dhK3CodEeFyjpk0QriTDzyGIY+NNL8ISGek4vYJTFPLjjO4gyS5FmCOhwO0Ww24fP5pt5Yk0/t8XiQSqWQyWTg9/sl+67ZH85K1KPyFJpyR09c0z71Qs5zESOnMda5Ec4hGn4uzKwNiMfjwhjodDooFApi9DWNlBzgYDCIUqkkzBH9faZVgsGgLKSEHDmOdGpYH6FLwi3LQiaTQSKRQDwel0jVsizEYjFbApDYMTnTFC4I2m6wYpZ2IBqNyusstAEg+QxnlF+v1yUap51gVTYjM87lTqdzLNH1sYKrTpoVlVWzPChHwRzao3Yaaiftz3kccKjwmletQ3EdDjGU0p6appER2uGkI1WPn6dB0N9vWoWLozEGmUwGkUhEKHJ8Vk4mDo9zGjpthDVNU3Nbb/fDsdP8e7048EfTMzXjiBh7s9mUilttrEOhEDKZDILBoNwf73naJRqNCseaVFu9eDpbTFA8Hg/i8ThisZhQ8WjIAQjljou1js51spk/2uHjWGsutXYGAQivmvdM75vziPfCKttGoyFjT/vB805aJm6sGSISgqDHojP/2rtxelKU23nUgB3b5nn18bcz+s7r6AdK40Fl5T1T4XldJhOCwaCs9MA49Nf4/LQLWwI88sgj8Pv9wtLRCUA9xhrDpnBstSI5aZVOh4BGlkaCrxljhGLFUncu4oQ2CNHoUFsXUzDzT+w6lUrh1KlTePXVV0WhdVn8NMvi4iL29/fRaDTkeREGoRF2EgZGoxHi8TgSiYRAKLowibpFT5z6qgkNmmZJ3dUePlsYOJlpwGFPm2w2i2q1imKxiP39fXQ6HYHQ6J2zyKter4vBNsYIrApM3mBPfEYRQnDizQxlNIakH5oeOG2gtaHUnwNgU1L9mvaqnOJMcgGwXYP3Rq+A3FtgbJDr9TqAMQzy1FNPCQY/6Wqlkyyc8OfOncPa2hpqtZp4w3rcCVnpcTjKAGvqncaiObeY/GX5MbndOv9B422MQSQSQa1WEyNMds9gMBDcWVMISTlkIooc/na7jUwmg8XFRaTTaQmJZzAIRE80DKrLt3VCn1WtLJzSDAxGszpRrT1m4JDGSSqtk0XEhVPbCu0o0lvmQuv3+7G0tIREIoFsNovNzU0bnNLv97G7uyuRdCAQsPVBosNK731SMnFj3Ww2hUmhH4g2wkdNZGcI7EwqauzyXkQfpw0Er6FF3y8/x2wzB5xVUOFwGEtLS7YKqGlngVB6vZ4013G73Wi1WreU+FP5WEgEHHrFFB3eUqmZ56AB1jxY5gqcXi29J54nGAyiXq8LpKLnFw0ucLhI8HOdTgfD4RClUgmtVgvXr19HPp+Hx+PB/Pw8rly5IlHltEMhfGZMwDsjWkKMpEEGAgHh5PP5ET+u1+vyOcIPzIdpxoc20pqCexQ0oR06Egd4nGb8xONxdDodcd64wPNYY4w4arxn3suki+OOjQ0CwPYgtfEF7K1QnYbaiSnq1yhHQQ067D3qvIBdkTmQ+lw0KoRByJ1mooOcz9FoJH1QuLrSY5h2GQ6H0ntYZ9Y19MFnzuepk746KawjLIagwGHRFd/XSR6+71yoOZ8YRhP7JIzGhDErz8i5JjRG7mylUsH+/j729vbkuy0uLuKNN96wJcqmWfQCp7tVAodUOD4nMqjI2uHYdrtd5PN51Ot1GGMEqgoGgwgGg9JrhvOEbXAJwR4Foepx0XOLsIoz6gsEAkgmkyiXy7ZjeE3dr0gba23rJiUTN9Zs8g6Mm7kcxafWHtTbfaGjkkBO0V6ZDmuOSjQ674EK7Fw4+JsPv9vtotVqyaAAYw87EonYqiGdi8W0SjQaxcLCAlKpFGq1mq2ikd40YQlik3ry67BWz4FWqyVcWs27tSwLzWYTlUpFME02A7IsSzrrURFrtZp44bVaDZVKRbBGFskQ3+Riox0Hhr/xeBw7Ozsw5rDdAaOBaW85wMgFGBtnFqoRs6bxHgwGmJubQzablUWy1WphY2MDm5ubKBaLuHDhApLJpLSgaDQaqNfrqNfrYrBZA6EXZ1JqNS2ThtvpQFiWJa0jyPrg684Gbex3QuyazgLPScaKs0XCO5WJG2tW+gGwPQidcNTJQeBW43Y3kIem5zlfA44G9/V7NOBHYWAMsXSptC6TpVfAklaGcdNupCnaa2o0GjbsTisUKVFOr5rGnJl5hsw0+pou2W63xWgzgUjmDj1mr9crER8pd9phYJ6C19/d3ZXIIJVKSdg+HI53FiqVSmg0GlhbW5OWnAAE/gNmjBCOE/tnUDjOmuY6NzcnPWNcrnGfFhrf7/qu70I4HEaxWMRbb72FbreLtbU1JBIJ9Ho91Ot1wcQZ/ViWJXknEgBCoZBUpnLHGr0JCcebizzHXOs5cGsPfK/XKywnQnGsH9CJ6knIxI11LBaztS6kAvBvwM6N1vQrwG7gtTjxJqf3oo03B4+YEq9zlGhDzQXE5XJJ43N27eKKC9y6O4mupJwZbPsz1d3wdLiooS7+7cSWNe6n6VU6AQgc4oOa56qVitgmP0djTS41oQ4qLsN39qLQGX6G9UyKpVIptFot1Gq1qTfQWqgrmg2mnSt6pZqmp8ePxjKVSqFcLmNvbw83b94UmCoYDErilxEacxo6H6LnAXeuYdUq4ZZmsykLSzQalVyLTmjqYjqdIPV6vWi1WjYEgM7KpG3BxI318vKyKBMJ67oeX3tQToMK3MoGOep14Nb+xlqcCR5n4uoow09PTffMZYMfGmWuuAyp2HBGG5cZZj0WbaTZL4YZd40fOg07lYiGlOei96Kb7nC8GL4SruL+e7wmvW5tMHRrgdFoJAae98SmTJFIRCAU3VuC3vjZs2exvb2NUql0S/5jmoUdKjWTR+O7wDiBR/iD7YZpcAln9Xo9vPHGG7h27Zp07uMmBHNzc1J8xXYA1EFSLLUDRo96OBxibm5OvPFWqyV7L8ZiMVsZOxd14NAp4HzQxAMuSnTeOIcnKRM31qdOnUKpVEK/38fS0hJef/119Ho9VCoVUUR6ScSCNE1HMwaIFzOs0J6VXr3o9dKg6kywMympDQMTSjrT6/f7EQqF0Gg0UCgUZPPd+fl58QAYvi0uLgqWSqMw7ZQt4BCjbDabUorLQgTu0MIwmXQ5HY0xPNV76KVSKTQaDRv2rfuwsH86IRGXyyXKG4lEbMZbtx+gwlEB9WIyGAykJzLD6m63C7/fj1arhc997nN44YUXMD8/L9AMcGiUplnm5ubE8LLSTxvVUCiEZDKJtbU1W/+d4XCIdDqNUCiETqeDt956C4PBAOl0WihyiURC9I7PudVq4erVq1hYWJANunW+o9PpoFwuI5FIYH5+Hn6/H//1v/5XXL9+HdVqFU8++aQwly5duoSnnnoKsVhMbAwNtc/nk3JynhewM0HoxD30MMjLL7+M3d1dwaRYg08jfdROIFRUGlDd4EfzXbV3rXmzLGrweDxSxqqzz85ssG6+xLBXZ6g1Z5O0IWMM5ufnxVgDwO7urhhpJkdmoTBk5/JmsynPEoCEqhwLjpFWKhpUvTWbZVni0XIsyTSiQSgWi7dUkNIQc17oDo3aYSCsRQXXRTPGGOzu7qLVaqFarSKRSAgbJJ/P48qVK5ibm0Mmk5HQeUbdGwv1luOpK4IDgQCi0SiSySQACKzJKImJw+FwiIWFBWmTSsgKgGy4zN2ber0eMpnMLdEycyNcjEulkmyMQFikUCgIZs4WpyyF15AHaanRaBTBYBDFYhGAvfEYAEmkTlImbqzfeOMN7O3tSYtJhsEAbCEKcBgGayzrKBhEU3F06HxU+MzQFrA3g9cUQL0QaMU+arcXevcsN+ZCUy6X8dZbbyGVSiGVStmOmXZFpTdFLI9esm5Tyl4bgJ3mpOEx/s/F2cl71h783t6eLdrieOpx4bmoVJxzGrqiV6yVrtlsyrZUkUhEMO1Go4F8Po9kMolEIiHh77SPP3BoqHUugosZ8WrmhDRUwQWVniuplkzuErLQsJVe6J1Qq/7huHLhDQQCSKfTNv623++XBDedOt0SlRBoJBJBOBwWGFdfkwsUG0xNSiZurF999VV5iGwjypWIKxtg72nrVCi+T2+H+JEW7RlT+KCcRHltqJ0eOh8+rwfY22+SXcBJ5nKN221+5Stfwde+9jW88MILOH36tI3pMpMxvalUKsnz7/V6tpayZGwAENiCVDld7MKqQSovt1Tr9/uo1+sol8vyQ2+Hcy0SiUiCmPfQ6/WkQRBgL57Q+RDdMoEhOT8TDocRi8Wwu7sr3reGz2g4pllYUaqTbtTzfr+PaDQqcAJzD1xcdeGUZv7QcQIgtoHeNeFK3aaYToC2CdTP+fl5pNNpgbdY2ObxeDA3Nyf5qWaziXA4bOst43K5kE6n0Wq1cOXKFZvOa8qerqKchEx8RpXLZaFJnTp1SsIcNl7R7AkaTe2N6J2DgUNvSHNdtSfGFY3FFbrj3VGsEs0woNdPo18ulyWZyPOSwgcAuVwOsVgMg8EAb731FtLptHgJ9Px4jWmXcrmMzc1NXLlyBfPz8wJnDAbjznYcz2w2K9xpKhQ9KHpVTPhVq1Ux1oQiuNv83NwcEomEJI3YfAmA0K+A8Zg3Gg2Ew2HxpAhrUcH1HHK5XIjH4/LZ9fV1pNNp6XHNLcs0tKJht2mVvb090Q12XKS3Sn50OByWnJGTI82FUu+jyTlEjvNwOJTWtRwzLuRMKgOweeDA4XzQlZAcN0ZtuklbpVKx0XddLhcSiQTa7TaSySTy+bzoP8+n4b9JybH0BqGnxBvXPUK0aOjjqJJQemFcVelxaSyKCUZN8dI0OmJkWpzhtBMfp2LyfKzAYo9j9gCYn5+X0OuonhbTKlwEq9Uqrly5gsFgIHQr3W+43+8jl8tJCEqlo3IR6uBnS6WSKIGmYDGkZs90espUQN2ljQu6k2cN3NqxkT+8D94j56vH40EikUAgELBFhNM+/sDh5tc6qU/niYk60ul0ZK09cOCQksvnS6cMGLNJksmk5JFobOkt0/7cDp7Szh+FC4QeQyekQzw6Ho9jcXERxWJRkuXcfcrpOE5CJm6s+YVoxPg/36MwoQfcilMzbGGnNnpcfCCaDkTDEA6H5QHTINyuQEZj3cRRqXz9fv8W408+J38YjnG/OM3J5D1NszCUbTQaWF9fh8vlwtLSEhYWFhAOh21YXqFQkMRzPB5HMpm0PXuOExvGs/qMSUBCbMYYFItF8agYDlPxCadoqEuXFlOcNEy32y2VuMYYpNNp8eYjkQgSiYR4jjMIzC7UA23sGC0xx9TtdsWo6RwW/z/K6SL1kpvSauNKBgYdLOd8o9Ax046ivpaOvp1ccQACvSwuLuLNN9+Uc7IVLJlpk5RjKYrRjU7K5bJwksm31Z24NE2Pxpl/c7ukwWAgpcNUvkwmg2QyiXA4jMuXL6PRaAg3mp3ddPZZe0xUeLIViK8zFJufn0cmk8G1a9ckMREMBiUEp1KeOXMGS0tLsoOIpvhMs3CBsywLlUoFr7/+ujR0X15etnGeM5kMqtUqGo0GSqUSarWa9H7Q5cH8zQWz0WjI7iI+nw+vv/66ZPzZunZhYQHRaBRer1cwZ7/fL5shUBE195d8amKoLIhhQ6KFhQVRxtXVVYTDYTlGNyyadKnxSRM6PpomS6Pn7HJIo6aZX1xM2ViLRlLPHQ1Psd6BezZ6PB6hXTKBqHnYjLw0FZSvM0dGp1FDtcxnWJYlnj2dU1bq6l1mJvpMJ3o2AGfPnsX6+roUJ5RKJUkchUIhMZxOXFd71k78moqkHyonAYsXaBz0Cs5BdzaRofAh6xXT7XbL/nvhcBjpdFp4k4lEAtFoVBajxx57DMlkUu6BK/O0e1hOj2gwGCCfz8OyLDz22GOSg+h0OgIhcBukRqMh0Qs9Lu6vVyqVUC6Xpeud5tp6PB4sLy8LJh6NRmX/xGg0KokqHb1RiZ2JYT0XqfhceJhwmpubw/nz54UdwrJzGoFpj65orOltUodpnLUd0HAkYY5wOCwLoYYXWV4OjLf1YjGNy+USvWTV6f7+vkBgLBfXcAsAW36Kog3zUVuT8d7pVOqx1lHZxJ/ppE94+vRpW9NxjQcfNYH1QDiNp6bW6XJgHqfLim9X4qkfnJ4gxKaoYLwmQyYaC91Jj14aExTZbFa2BNIY6bTLUQsWi1aKxaLsq0cuNvMAurUuFSESiQhHmp42WR1UJjZdYoYeAFKplCQuyZHV4+9UqKMWWk0J4zxgpaLX60W1WhVql9frRTKZlGhtlmQ+NIx8rpq6q40mqZnAuGsnYSxtDwD7jk9MVCYSCdHZWq0mvGfu5cj5wvyWpv9q2FLbDSeOre9ZU0oHg3EDKb2jubZzDz0b5MKFC7h8+TJ2d3clLHQaMo0JAbAZT4YZzgemGSE62adDVp7HSZsipqWVVXf90vdEGh/Pyb4BDM352WQyKV43AFsJ9LQbbO19UBFItdve3kYqlRKcd39/X3bjYFUok9LVahWZTAbRaBQul0sSkKwgjEQiyGQyUllKz9zv92Nubk4WW+Y0dG6BERhgj7D0dlE0xFwouJN1u91GpVJBuVxGJBLBysoKIpEIVldXsb6+PvVMEOBQr/hstZPk5Cbr5D73Z+Sx3ESChpHU2Xq9jvX1dSwvLwubp1Kp4NVXX4XL5cLKygrOnDkjRXn02LUdol3SBXb63vk/bQMjAO7Z2e12Ua/XbQiAxuAnDYVN3Fi/5z3vwVe+8hWsr6+LknLAWGlI0VgQvRMODGDP1jIsHY1G0ugFOKx+1Fv96B8aXgoxL1KGeD3NMqGhZvk5MGYj6Java2tryGQysp2XTo5MO15JL4p/89mMRiNcvXoVTz/9NNLpNNxut0AULC1fWFiQhTybzUofiMFggPn5eXQ6Hfj9fgyHQywtLWF5eRmpVEpeZxe3SqUiYTh7T+gKRm2UAch7nK+ahcD79Pv9yOfzaDQaAsOUy2Uph69UKhINTvuCrYXeJj1PwopkUenkLKmUfJ/6Rz0OBoOSe0ilUhgOh9Ky1O/349FHH5WIm8lMwlncIMSZq+j1eiiVSjYbw2trPJ33oRcXnQMjrKcdgUnKscAgZ8+excbGBm7cuIFsNithLQfGiSMDtzZbAg6NNY/XYYiGLDiQ2lgTFtEVTeQ/8jiNfWuIRYe97AOSSCSkw5rL5cIzzzyDSCQCj8cjpPqZjOUoQ8WFjMnERCIhPSD6/b5sPDs/Py/jTA+cEgwGsbCwIMa61+tJv2yOL+eQkyJIzFEXyOg5Re+JCq7nhm59urKygmq1KuXvbJpPPPW4kksnTfj8mGcik4u6rCNuPm/qMfnP1GdN2dUNuEjX49zSFZHURxpo7lRP2EwzUnTUznsg7MF7dnrexMLJ6ed8YB+aE2Gsk8kkzp07h5s3b+LatWuIRCKy0gG3brOlk4PAISRC7+woxdcYs6ZYaaOtz+18jUqqqX16QdAccfYwSKfTsn293+/H448/blv19feY4ZW34pV8PuwnPBgMBG9cXFyUzD5LePmc9Rwgr5nsjnK5LH0hwuGw4NiAnZpFQ8z3qGg6Gc15Qq9KR3P8HsB4weCxg8FhgyG2JNCc7WkWLnQcA73Zsabd6rGh3tFYay+YY6UpdZrf7nK5hMfPOVOtVkX3vV4vms3mLcwP3ivH2pn/4v3qxVezSNi1jwlszemetME+lprYD3zgAzDG4LOf/awMBj0abdTo4WrSvDawDEO9Xq/N2AOwDbAzqUjMmwNDLJqdsHTYxfc1fMH+EF6vVxJXq6ur0oQ8m83i/e9//y2eGO93ZqwPxbmA1et1FItFLC8vIxKJSHERn53eKZqeEHA4R9i+Uoeyw+F4WzXCVLwOcybcZaTdbkvim8cTr9SJRF5Pe+aca/v7+2g2m0IzPXv2LGq1mnRvI15JT35ahQsiqa66a6aGDZ3QZSgUkryE3kRC96nWDBAmkDluerGMRCK3OHNMZurmcdoYU+91DYh2DAHYFp1WqyVeNc/PSGvSC/bEjXW73cbS0hKee+45fOxjH8OLL74og0CuKwBbkkErNOENKgqVVK+ugJ3l4ewRQg+Z/OhYLCZ9aiuVijxkTeMiRqW9ap/Ph1gshvn5eaysrKDT6WB+fh5LS0vSBwI4pAHpqsppF51x19CVy+VCpVKRlgSJREI8E056AELf0mPvcrkkXA2FQtKaFIB4VSymAoB0Oi1KzLGighLPZkk0cKjoNOY0JCw/9vl8+I7v+A7bIrCwsIA/+7M/w9bWlszfoxyIaRMNJRL6pH4QlhoOhza2Dn/0Zh+6hbE2iDwGgBRJOb1iOlP05Dk/uIho4gPbYDgbQQH2HYAA2Dxwnp+LQavVkvt/6D1rGslkMon3ve99eOWVV2Q11AwMPnSKVo6jKht1ApDCVdfZMAY4XCH50Lia0svnA9f0QieFBzisdKPhByB9DrQHrUO6mdirOPVEp7FmUyfdS5zjwMnuPJ+GREajkfT3YFSmvSxGOzTOwGGTHV6LUAaNCpkJeh5oqGQ0GklSmQs5S92dEcC0wyCMjmno6MToZD5LsolBawyZ5+B40bPWtoEGVXu/Ou/ktBXaM9bn0N63E2vXETPH1MkoIVNEs030vU5KJm6sOckDgQAef/xxmcTEl1gmDEDCWU3b09Q7nWBk1lU/TJ2F1R4xAFEoekY8nhODnyG1h6sjlVVPBCYgAoGAbRsg5y4xM7GLc5JTSHtjG1Wd4NVJJnpQeuwp9HY1bKEXaLJ5OE84tiy80AaYBpv/6+SkrnBjYRf7nDDc1UlGGoVpN9Y0fMRwgUPbwMWTETdpdTqa5g+dMTpbWjgmTpiDNkUnBqn35HNru0LaLw0zFwhS8vTirfFt6r2zXkM/g0nKxI01GRcsKw0EAqjX61KzXygU4PP5kM1mbVlz3eSbDxo49MqIedEb0j8MmUmmZ/hE8B+AdGCjB0aGh8bDaKg5EOT6chJks1lpGqMNx3EkE066aGOnxbIsFAoF7O7uIp/P2/oWu1zjHiI8ToeXxDmZba/VaiiVSjDGyBZPOovPcabysR0A54tOJmuGgs6T0BiQFka4pNFoiIFJJBKCr3OhmRlryOLHAiIugqFQSKJbet6kUfL5AYfGlAVQPCehh8Fg3EyNzA6OPWmBhK10klN7+npeDodDcQ7IMGq1WohGo9LJkTaJx+nmYLFYTOweSQlMlk/0mU70bDjMlLpcLszNzeEjH/kIXnzxRenfwSQdu5c5mRQ0etqr5mqpPSFdwsoBJo2mVquJlxWJRGRB4Pmq1SrK5bLsYk3RBpf3wqY9LMB56qmnpB8I8VKu2Dp7PO3i5BrrxPFgMEC5XMb169fx1FNP2fqJs7iI0RAXZHZR475+wHifv3K5jEKhIBAFz0+lZnKLHhU3OKWi8dqcVzTkOrymwWHIThkOh0LXc7lc8v5s4YYkj4fDIeLxuDAxqBv8rZPzdNac1EmORafTkU0DyIHWOhoOh+V4Qpb6ms5x0frKqJlzjVAZPWxte8husSwLiUQC2WxWdp+h3WE/pEnKsbBBNGPjqaeeAjAevHK5jOXlZYxGI5RKJeTzeRsGBeAWY60xayfmqA09FXwwGKBWqwlxnlxY4LDdKRkHhFw4mLpMWGPavD5Xbe7Npr1r5/eYZtGQk07i8H9WM25sbOCJJ54QI833mFTSeQ6ej5623+9HMBhEvV5Hq9XC9va2NF7SnjGP1ZWlTshKJ721YnMBpmdFT41jTU+K98PvDcw6L6ZSKVSrVfT7fUkcUte4QFI0FRc4hFBoSGkc2+02crmcFMdkMhmbJ3zmzBnp08PGbjoH5pwDGhvXvwmNaJqgzkXoHFcmk5GinEnvueiUY8GsNRXvqaeewurqKp577jkUCgU8/fTT2Nrawu///u8jl8vZKol4vF7BNAGdf2tOo27iwj7H2pjrfd64arIPcigUumWzW50c0NgUMKaDcaLQcPAzOrkw7YrKkJHPgnMBgGCYlUoFV65cwbd927dJVp+90LkLNcNmRleE1vh6NBpFqVRCr9fDzZs3EQ6HkUgk4HYfds6jUtIQO5NZgN1YO++Z0AsXFN3YSbfJZTn7jA00lrW1NRSLRWmupPNEpNRRX3QbACeLiGPf7XaFIknH6dy5cwgGg9jZ2UGlUsH8/Lwwi7rdLqLRKFKplECjPJfGpoFDR4tGmpsnE0bTuRdCcsTQl5eXUSgUUKvVUK/XARwyjhi9TUqOxbPmQ6Hhi0QiiMfjOHfuHAKBAKrVKhYXF23ZVnosuiWhDkX0LudcrfVgAmNDEA6HbRg0E4haGKKwWYwTA2fiiiEQ8TcaanpvvO4Mo7SLLs11FkDoiGV3dxcvvfQSnn76aWQyGbTbbZn0vV4Pp06dktA4HA7DsixZjPv9PsrlMqrVKizLwu7urmyvxgY+9Ih1kYtmDfEzbL9Lo8BiJx0267nE+dTr9bC0tIR3vetd0lHyC1/4AorF4sR3CTlp8kM/9ENisDTu7HKNd1mp1Wool8uyHycjVeo17QDHfzgcSgtd0jy/+MUvyka5g8EAf/zHfyyLwdzcHJaXl7G2tob5+XnZKEQTBzQLRCcSiVNrZof+m/fMntaPP/44FhcXUSqVsL6+jnw+b9upaFIycWOtMSaKxo1oCJPJpC2TT+OqjbceaBppZnh5nKb0afiFg3A7TIxGmPcEHEIvurcHMVAA0naTyk1xemgzw32r6DGjjEbjXiErKyuIRqMYDscb4JKDS270YDBAu92W9qj1eh31eh3hcFg2s9X0MOLSTk+Z0mg0xEizwMKZcKJwYWm322i321JIw8V7OBxKy9RAIIDnnnsO6+vr+MpXvnLcj/ShlnPnzuHpp59Go9HAG2+8gVQqZevXUa1WkcvlsLm5ifPnz8tY8dlruIuFNW63W+YJP6MZGQBsBrVYLGI0OtxHkfCljoJpE7R90Eln7RBqCFbvfM6++ozgyXSiXZmUHMtOMU5D5qTieDwexONxW2kmYMeUtIHmb0IjPE5T+ZxVRpouRIVlhlj3ktCD5gx/6VHzHDTWTpI8r+PEvqZVjlqsnPAQn2kul0O5XEY2mxXPlt4vDaTGK/f29sRYcwswRlocQ46b5tBr2KzZbEooTs69Fu00AIeZf52/0N8rEAggHo/D7Xbj3Llz8Hg82NjYOKanezIkmUzi0UcfRaPRkEI0whMAhGufz+dx4cIFm/eqnR+9ywsXWHrC7MgH2Jkk1GtGaMPhUCqQmcvQdRGcL06GF422/iG7hOfhPRE+SaVStjbLk5SJG2uNNQH2UnIaa6/XK54IvShn9pcZWT4g3WhFe2mj0cjWUU0LvSm9Hxs9KXrfznPTk+r1eohGo7Knmt/vt/Ww0LuCAIfGx8mCmMnbSz6fx87ODhKJBBKJBDqdjmw2MBgMsLu7Kxvl1ut15PN5tNtteL1e8bwBCPeZbBKOLT1vl2vcbjMYDKLdbkumnuPtTIBqLNVZ8aj5u2SBEB5JJpPS9Gva5dlnn8XCwgKCwSCuX7+Oer2OTqcjCcFut4tcLndLbkgzQpjk1UKPO5VKCezFeg6em4wfVsXeuHEDiURCcgtcqGk3mC/RdR/A2MboHas4/pqlRpvFPEkoFEK1WkWlUpno85y4sdYEcW2EgUPviruCpNNp5PN5wYCAQ8aGE4/WW/KwCZBzEPXKx+uRQ0nqD7mZPJdu/tJoNG5Z4bli+v1+PPbYY4JnaS9ef7eZ3EqRohwVjQyHQ2xubiIYDOKJJ55AMBgUpofX68X8/Dyi0SgajYYkcogHk3bpdruxtraGWCwGAJJA1kUYbEZP7JsQBj13nWTUyW5CHjTeLEXnPeqF3+PxYGdnB6urq/ju7/7uyT/YEySj0bglwJkzZ/DDP/zDguFWq1W8/PLL+PSnP439/X1sb2/bdluhh8zx0waUMEiv1xM2CMeVyUAWyZFwQGPP4ivNUHLy6nnfOirXCWrahDfffBPpdBrZbFa8e+bY6HC2Wi3s7+9P9JkeCxvkqKofneVlCJPJZGTHaid8oalxOiNLRWIfYXrLWsFYvMCSYM1/9Hg8qFarACAN8Mko0Nimrj7y+/3IZrNYXl6Wa2mMVH/3WbLxzpi9ft/lcqFcLiOXy+GRRx6xhaYcPyaDKfTQGFFxswFNtQIg78ViMWGRALDxZhnZcc5qmp+eixxvRnsApLeyZpvoXdenWTQNjwY1FosJrnvp0iXUajWBtciR1gk9nocLISNxjiW37+ICz97yjKS9Xq84fgBs+z5q+6R1ljaHhl2PLYkO3N6tWq1KxAaMI/jLly9jY2NDIsBJyrGwQWhotZHmSgZAMqmZTAbXr1+XB3jUOYDDun79AOv1uuBFHEh6yPS6Q6EQTp06hRs3btiqFq9du4Z+v49QKISbN2+iXC5LxZLGsngPoVAIy8vLmJubkz37AHtjfX4v/Xta5Shjfbuowxgj2CWb4GiIgZ4wfxMqKRaLMo9oCOglEytkZKe3Y+NcogfEOcVx5LHam2K0aFmH3ePIDOICT0WmV8XK2WkVJyyo+/IsLCxIq9tGoyGbarMuQrN2eC7CC3obLRp4RjnUTUZNOk8GwBZ164Wd96rv3Un35Hxh7/pCoYB8Po/hcIi5uTnx3l955RXcuHEDzWZz4k7bsXjW+iHrBB2BfHq4a2trePXVV20PlgOjiw9oeDUbhD2NueXWt37rtyKTycAYg3e9611ot9vw+/144okncP36dQBjyl46ncaNGzdQLpdRLBZx+fJloRVFo1Fb2XA6nYYxBqurq/jIRz4i5ekk3vPeKDMo5FZxRh4U7bnSW1lfX8dTTz0FY4zwmXX0RDiDSUJ6UWx9ymIbejWj0bjkfH5+HqlUCpZlyVZMTtYBvS/dEwKwt9xst9soFAqIRqMSHTL52Gw2USgUkEwm0ev1bHj6NAuNJZPA7XYbv/u7vysYNgC8+eabSCaTSCaTqFQqUqbNxZHRssvlQqFQECg1FouJPdG2wuPxIBKJSIsBLhIsXCK2rKNjZ9IYOIRVdTHecDjEE088gRdffBFXrlzB5z73OUmOE4+nw/HQG+ujvMqjmBqj0QhLS0uIxWJwu91Ci9I0GX08Q1hyZpPJpK1vxJUrV7C9vQ1jjGzxxITCxsaGrLilUglXrlzBzZs3sb6+jn6/j/n5eRt0w7Cr2+3ikUcewcWLF/H0009LGMWiGf3dgEMWwe0w22mWtzPaLtd4X72vf/3reOyxx8QzKpfLAnkw+mE5eKFQsBXfcPyoMKTUceEn9MXFlopE4+y8Hx3ZMUzudDqoVCpIp9NiKJjE2t/fR71eRzAYlOKsaRYNexLj52Kay+VsW/NdunQJZ8+eRSKRsB3P56uduHQ6LcncVqslvUZoB4gb66ieBXBMKpM0oA22rpikkWa0xNfpWZM+SoeC/YOMMZIYP47iqGPruscHpVcYPhgK+Ylutxv1et2GBfPzPCexbuCQwsMHTgVk2MuEVbVaxXA4RD6fl4qjZDKJra0tFAoF2UlZ9xah4hMzffzxx3HhwgXxonhvGiPXmDrvb5rFOebOKMSZ/eezK5fL2Nvbw/z8PAKBgPCnG40G8vk8gMPqSHplzOjrJA/3c+QPebr0oOhxafomFVYv2Dr/QgeCSWd29eOmysViUbYrY0uDmYxFw4OEM4DD6Gpvbw+lUkn2aNXwAw0ldYt7dtK5A+yUX44TPXLgsFjO6SxyTmh7wx+eh+cm179arUrRExcB3YjK6YhMUo4Fs6ZoY80vrEOPeDwujZZqtZokBumh6m5bGkvW4bHP50MymbSt5Ow54na7cf36dVtSiNxcyxrzpknl4uBqJc5kMvjgBz+IRx55BABkJdXVTRTdz2DaE4xOeTt4SGfnB4MB3nrrLbjdbqyurqJQKAhsUSqVpH80efps3hOJRKQLG5kibBjm8Xhk+y9GZfSEuGO5VlKdr3BGTf1+XxJaDOlDoRBqtRq2t7eluo4e+DSLXqD53H0+H+LxOE6fPo033nhDjHKxWEQ+n0elUsHKyoot76CTvsYY2emeDdvo8XIRZf6g3W7bdJlMIYo2yroqUePVetFgEyly/dn6gLQ/GvZJc6u1TNxYkyrDH729kea/sj806/dffvllyeJSUWj0nJCIcwXUOKM+HhgPCql7XKV5Dh7LxUMvBMlkEj/xEz+BRx991NZDhIOoq954XobOk65cOulyp3BQe+GXLl0SnjVb0jLxxyQjP8/ttdrtNkqlkmTp2+020um05B94fSpVp9OR3crZTY/3qatZuYhoT3ttbQ3BYFAMCvtVvPzyy6hUKpJHmXnWh6JbDng8Hpw/f162WiMEuru7i52dHaytrdkcJ0ZF1FWeh31aNEWTtoU6TpZJJBKRnAf1U7cgoLHX1dDNZtNWBLW9vY3d3V3s7e2hXC4L5NLr9QRjJ9xG6GTSRIOJWxV+YQ136Ayv/rtQKGB/fx+VSkV6AugtlbSiOUMMXouK7AT0NU8TuBWC0TxuGmtSg1ZWVvDUU0/hySeflE5rFJ18cHrQPM8Ms7470bAIcFigsr+/jxs3buDMmTNiNN1uN8rlssyhdrst3c7oyTLSSiQSwhIh35+sjtFoJFRAwhmapcQkuHNhDwQCyGaz0gOEVbXtdls2hybWzj42MxmLdt7cbjeWlpYwNzeHYrEo3vJwOEQul5NueVwo9QYS9HCBwx3r6bFTT1kMQwPMhl805IS9nAVttCMam6YUi0UUi0XUajUAEATA5XIhlUpJi4TBYIBYLCZR4kOPWevyXgoNI5MA5Czu7e2hWCwKhqgbQOnmKlQijTFpz8fpeQP20NuJn+oQR9+nMeMmLufOncOzzz4rzYXIw9WetP7txKlmxvruRUdKNKyVSgW7u7s4c+aMTHo2qWfRQ7lcRqVSkfnECIoN49mKk+OmnQSd79Dzypl/4LgTZ02lUnKPNNalUgm5XA67u7u25mLO5mHTLNrRIrV2cXFRGjmFQiFUKhVp1ETyAGDXJ3rTwCENT+eRAMhOVFxgQ6GQNPbSToHTCXQ6dZrFxmIsLg6cr9z0gj1I+v2+0Ag19DopmbixJg8RgPAodYMUY4xQoF5//XXs7e3ZMsP80nwYuu2ks2JQ41q6TNjpmeswTH9OZ41JE3rf+96HF154AR/84Adtezvy2noAnJNQY10zubPoxY6KwUrSXC4nE55hayKREKiDCel4PC6FM3rXkGazKV4zKXv0uNjdkfx+tmHVSSodQSWTSWSzWSSTSZsz0m63sbW1JQnrRCIhOOYsb3EoTiaQz+fDM888g36/jxs3bkhEWywWsbe3J4UzAIQ7zT4hzWZTdNHr9aJcLgt8St30+XzSXIvtbckAYcSl7QqLZ9jgiTsDjUbjJlA3btyQvBbZRqyiXllZQSKRwPb2NnK5HABIleWkOy8ei7HmwySfsdvtolQq4fLly9jc3JTWluQmMlzVO6/0ej3pv8GHqHFpjWfTe9JeuTaYGk+mdw/YGSuBQADz8/P42Mc+huXlZRtPl+GYk9UA2BunOyvfplWcTA/9uvP9o55Vv98XBgjbnrZaLcTjcSwsLGBxcRGPPfaYLQ/BBRiAeM96XDQspneiASCYNecj78sYg7m5OSwsLEiCiqyiXq+HQqGAL37xi7h69arsgkSmAtkq0yq3SyrTmTl//jzcbjc6nQ7+8A//EMBYl65fv47Tp0+LrnNRZdKXrA5jDGKxmHjSOklIOIztANhWgl0WA4GA1Gk4DTYjuXA4jO3tbayvr6NQKCAej0tJOu+p1+vh1VdflYgrFosJNEebM0k5lt4gxBUbjQa2trZQKpVQKpWwu7uLjY0Nqe4iJ9EYg0gkIhOcD1yHuBx4vbEuXyPE4mx1SeGD0yGuput4vV5kMhl85CMfwcrKCoLBoHjUNOxcAG7Ho9YJz2k31vciThgEOIyarl27hmQyKUUo9IgA+56cAG5RDv23pmHpvAnngjPSGo1GUoiTTCZv8Zb5mUajgRs3bkjOhXxt4qnTLE4Y0mm4+WxPnz4tW2oNh0Ps7e2hUChIGTfZNdyqiwaTHrre3Z5zgfaBxTIaTtWsLQrxat1Rj9HdxsaGGHdjDlu2Ev/OZrPS2I0JTF0hO0k5FtoCs+F7e3v42te+Jm0wu90uNjY2YIxBKpWSpICur+cDJ13vKGxZc64B+3ZcLJqg0BgcZcTpdaVSKaytreHDH/4w4vG4KDCLNThItxOnxzgz1m/PAHEqsX5eGn5YX1/HmTNnRFEJhzDhxMWciqGvqSsQaTzpEZGCyfc1J5fn4oan0WjU1ned98tocW9vT/pVkHVETHua5Xbjr6NTnbTt9XpotVoolUrY2toSuMJZMcwo2+U67JoHHM3d19G001hrw67pgTx3oVDAzs4O9vb2pLsfHTvOw0AggFOnTqHVaslWgYRrGMFNUiZurK9evYorV67g+vXreP311zEcDuWhbm5uYm9vT/jVhBcGgwGCwSCi0SgqlQrq9brwr/lgh8OhtFWkV8UHx5VMJym18ebKycHU7Q+j0Sj+8l/+y/jQhz6Exx9/HKVSSR66zh4z/KVow+xMVkw6C3zSxGmA9evArRsQaIXhOALjRX9rawvGGCwvL0sTeS7ietNk3bdhNBrJ5zhvGo2GfL7b7Uq3NEZPOtriXGS7AZ3bYJKr0Wjg+vXrUhxBo8FFZNp3ijlqrPm6puLpplf9fh/FYhF/8id/AsuysLCwgGg0Ki1K2UKCiWFuAADAFhWRIqijXMIU1Gd669qRI+xSLBbxpS99yRYxcRHnQs6mVIRAstks3G43rl69KtG/7sU9CZm4sf7DP/xDbG9vS/VgJBKRRj0bGxuIxWKSodXZfHq4VDqWkjoz9DrLyqbf3ICT2KQzKQgcbuKqk1nD4RDvec978N73vhfnz59Hs9kUz5yh8tvhTs5QSic2Z3Jn0ePKhS8UCiEejyOdTmNtbU2MKncw52Lb7Xal7aVmBnBMuOgOh0PhPvOazujHmbmPxWKCjeq8BpOKbNe6tbUlnHsmuHRBzUyOZkrROCcSCVy8eBF/6S/9JXzpS1/CaDTCI488go2NDXQ6HeRyOSwtLYmzxLGgTjrbIWs4jcZVkxvopXO8mddgJFStVrG/v49Lly6hUqkIw4hYeCgUQjKZlAQmWSbxeByBQACJRALRaBS1Wk02XZikTNxYv/XWW9Ioh3utlUol2eqGmw4QoNcPnIY3EAig0WhIYkHjmVw9mRygQdV8WW1gySYA7LsX68IcPnjuiq4Nuqb7HIVJOxkNzutPo9yO2uj0srTyud3jXsXz8/NIJBKIxWJCwWOvDS7AVCDt7bDdAMeH7SlJ2+v1eqK0mhbG44HxuGl8UjOLOKa8F+6qriMBOgQz2t7tRT9XcuKffPJJrK+vo9FoCJZNau/S0pKwK0ijdS6GzsWXi7ROLDO6ZiRPGwCM8x/1eh07OztS+MJ7pY1htSuraJm8pH2h7eIekGz9PEmZuLHe3NyUCrRwOIwrV67I7g0ul0tWIRLJdYjp9/uFxF6tVkXBNNZE75qNxXXFpG72RHEOpD4PS5mZHBgMBhLCkMWivWUnLnaUOD36aZXbLWj8Xxt0Gurz589LxSgA5HI5NJtN2Q6K3hU3OtYelrPkl0lsZ/KJ9D5NteQ883q9sk8fFV5XvLEYp9Fo2Lx63bNGl0pPuzjHXUcoepF+17vehZdeegm7u7toNptIp9Not9vY3NzExYsXkU6nJQHZbrdl7DXDh7rprL3QizPPQQPLBb7T6aBQKODatWvY3t5Gs9lELBYT+8KcWiqVEg64tgl0KMmxZgOxSTfzMrNk2ExmMpOZPPwycwFnMpOZzOQEyMxYz2QmM5nJCZCZsZ7JTGYykxMgM2M9k5nMZCYnQGbGeiYzmclMToDMjPVMZjKTmZwAmRnrmcxkJjM5ATIz1jOZyUxmcgJkZqxnMpOZzOQEyMxYz2QmM5nJCZCZsZ7JTGYykxMgD9RYG2N+zRjziw/yHmbyYGU2B2YymwN3J3c01saYG8aYtjGmYYzZM8b8qjEmcj9uznEff8UY8zljTMUYkzPGfMIYE72L454/uPeGMaZpjLHU/w1jzKn7cf8H9/KzxpjXjDF1Y8y6MeZn79e134k8RHPgBWPMyDF+P3oXxz1Mc+AXjDF9x/XP3q/rf6MymwOTE2PMzxhjrhtjasaYHWPMPzHG3LED6t161t9rWVYEwLsBvA/Azx9xA8eyRZiSOIBfBLAE4HEAKwD+1zsdZFnWZy3Lihzc/xMHLyf4mmVZG/zsffgOBsCPAEgC+C4AP2mM+cFjvuak5GGYAwCwo8YuYlnWv7nTAQ/ZHACATzq+w/X7cM1JyGwOTEY+BeDdlmXFADwJ4BkAP32ng+4JBrEsaxvAfz64AA5Wp79tjLkC4MrBa99jjPn6gQf8eWPM0zzeGPOsMeZrB57lJwEE7uHa/8GyrP9iWVbLsqwygE8A+NC93L9TDryc3zTG/DtjTA3AjzlDsoOVfEv9v2SM+S1jzP6Bd3zHh6y+wy9ZlvU1y7IGlmVdBvC77/Q73G95kHPgOOR+z4FvBpnNgXdsB65ZllXhqQCMAJy/03H3ZKyNMasAvhvAS+rl7wfwHICLxph3A/jXAP4mgDSAjwP4PWOM3xjjA/A7AH4dQArAbwD4a47zV4wx33KXt/NhAK/fy/3fRr4PwG8CSAD492/3QWOMC+NV8WUAywC+HcDPGGO+8+D9bzHGVO7mosYYA+B5TOY73Dd5CObA3EEYvn4QPoYn8LXu9xz4XmNMyRjzujHmv32H937fZTYH3vkcMMb80MHCUMDYs/74He9Q71121A+AGwAaACoAbgL4lwCCB+9ZAL5NffaXAfxDx/GXAXwEY+O6g4MNDw7e+zyAX7zTPRxxTx8FUAZw4R6PWzu4Z8/B/78A4M8dn/k1fU8AXgCwdfD3cwA2HJ//OQC/+g18h39wMNj+ez32fv88LHMAwAKAixg7GWcA/DmAj5+kOXBw/0sA3AA+CGAXwH/zoMd4NgcemB14BMA/BLBwp8/eLTbz/ZZl/dFt3ttUf58G8KPGmJ9Sr/kwnpwWgG3r4A4P5OZdXl/EGPMBAP8BwA9YlvXWvR5/hGze+SMipwEsOVZNN4DP3ssFjTE/iTF2/bxlWd17OfYBygOfA5Zl5QDkDv5dN8b8DwD+AGMP7p3IfZsDlmVdUv9+3hjzvwP4AQD/8R7u4UHJbA6MZSJ2AAAsy7pijHkd48XvY2/32UkA6fqhbwL4R5Zl/SPnh4wxHwGwbIwxaqBOAbh2txcyxjwL4PcA/A3Lsv7rO7hnLc59zZoAQur/BfX3JoB1y7Ie+UYvZoz5GwD+HoAPW5a1dafPnxC5b3PgiOtOYhvx+zoHjrj2N8NW6LM58I2LB8C5O31o0jzrTwD4W8aY58xYwmZMuYsC+AKAAYCfNsZ4jDEfA/D+uz2xMeZJAP8FwE9ZlvWpI97/BWPMn07gO3wdwHcbY1LGmAUAP6Pe+zKAmjHm7xpjgsYYtzHmSWPM++7mxMaYHwbwjwF81Do5DIB7leOcAy8YY04dnHcVwP+McZKW75+EOfB9xpjkwXd4P8YsgN+903EnTGZz4O2/w48bY+YO/r6IMYRyR+dzosbasqyvAvgJAP8cY0z5KoAfO3ivh7Gb/2MH7/11AL+tjzdjvuPztzn9/x1AFsD/YQ65kTo5twrgLybwNX4dYyz5BoBPA/gk37AsawjgewG8C8A6xsmBX8GYVihczrc59y9inHD5ivoO/2oC9/zQyDHPgXdjrOxNjHHO12CnPJ2EOfCDGD+TOoB/C+B/se6CenaSZDYH7jgHPgTgVWNME8AfHvz8/Tvd0DfN7ubGmK8D+HbLsooP+l5m8mBkNgdm8s08B75pjPVMZjKTmXwzy6yR00xmMpOZnACZGeuZzGQmMzkBMjPWM5nJTGZyAmTiDUt+67d+yzLGwOv1Ih6PIxodN8azLAsej0f+HgwG+Iu/+AsEAgEEAgGEQiH0ej2MRiMMh0OUy2U0Gg30ej0MBgNYlgWv1wu3283KH/j9foRCISSTSQSDQXi9Xng8HnQ6HXg8Hvh8PsRiMQDAaDRCr9dDo9HAaDSC1+tFIBCAZVnw+/3wer3yOd5fqVTC3t4ecrkcNjc3sbq6imw2i0QigWQyiXg8Do/HA4/Hg0gkguFwKD/vfve7vxm4s9+QbG1tWcPhEKPRCADg9XplXEejEdrttjynz3zmM/id3/kdvPnmm/D5fADG84Ofd7lc8hrFGCOv6b/1Z0ajEVwuF4wxMMbAsiy55mAwgDEGLpcLbrcbnK8ej0fmF68/HA5vub6+D2MMRqMRBoMB+v0+jDFwu93weDyoVCpTOwfe9a53WZFIBD6fD8PhED/wAz+ARCKBXq+HP/qjP8Ljjz+Os2fP4pFHHsErr7yCl156CZubm2g2mwiFQvD5fGIb/H6//J9IJNDv9+V5e71eRCIRhEIhuFwuZDIZxGIxxONxLC8v44033sD+/j7C4TA++tGPIpfL4c0338Trr7+ORx55BCsrK1hdXcX6+jo2NjbQ6/WwtraG8+fPo91uo1wuy/Eulwv9fh9vvfUWbt68iV6vh+XlZezv76PX62E4HCIQCCCXy2E0GiEUCuGXf/mXJzYHJm6sR6ORbQL3ej1RBCreaDTCaDRCv98XBaVyUOkCgQCMMTIwg8FAlInCAeS5qbherxderxc+nw8ul0uO53lHoxHcbrcoFRcBABgMBqLoXq8X4XBYBp9G3u/3yz24XC65Bq9PBZ9W4fhy4fN6vTbDpo1ms9lEp9NBv9+3jaFlWXC5XLZyW22AOZeOktu9pw0/z8frcD5yTnD8vV6vszzY9t34QyPtNPLTKhxTr9eLYDCIq1ev4oknnsDq6ire//73Y21tDX6/H9vb2/jCF76ASqUCj8eDtbU1AECz2US9Xker1YLf74ff70cgEEC324UxBsPhEK1WC7FYTBwzAOh2uxiNRggEAhgOh+j3++j3+zJmAODxeOD3+9Hv99Hr9WBZFtxut5yD5+n3+zJPOScGgwHW19dRqVRs80bbBp/Ph9FoJM7ppGTixtrp7fT7ffmi/HJ8j38DYyNJYw2MDbGe/PqBARBvyOfzyWepaABkgdCKxfPyHmmoPR6P3LP+7fF4xOsPhUJiLHgfPKfL5cJwOBSDz3NMqzQaDTHO+lloA8lxZeREA6ef/1HGmu/xN8dA/w3ANj583+ll69/8DP/nPKLnrb8H5xQ/q9/X322aJRqNircJAPl8HqdPn4bb7capU6cQjUZRr9dx48YNlEolhEIhxGIxzM/Po16vo9/vo16vy/NltNvv98XLZZSmPXDgUHf12HKe8T2PxyNjRb2mbeC8pBHmMW63G71eD/l8Xr6jz+eD3++XuRsMBhEMBsVeTVKOpW8rDScnvzbKTiPHz3BgGZr6fD7bl/X7/TbPi8e73W5ZCPg/z0NPjl4PvSVOAEIfAG4xEsDhChwMBhEKhVCr1eT76LDe6fkFAg+04+MDl5s3byISidiUSBtGPit6LsChoQQOoxXgVi+Wc4nPWs8JKjUjJ23Aj/K2j4JQtKHlvKG35PP5bll89Hzo9/sAgH6/fyRsMk3y/PPPY2dnB7VaTeCDXC6HcDiMTCaDS5cuYWtrC9evX8eTTz6JtbU1xONxDIdDXLp0CZ1OR/SZkTGN5+LiIpLJJLrdLm7evIlGowGfz4dsNotoNCrRc6/XE6+8Vquh2x234aEN4JgC47Gm3jabTRv8Gg6HJdLv9XpoNptYWlrC8vIyFhcXEYlEUCqV0Gw2kc1m0e120el0Jv5MJ26s6XlqfJkPWkMG3W4X7XYboVDIBm84Q1QaWa1YFCquU8kZhnC10+EvDYReNLSXrcNshlfExhuNhi3EHQ6H6Ha7sooOh8NbFplpFP1c6W2MRiN0u120Wq1boihGKpwjXGC1Mef5KNojOsoQawdBH+s04NrAOxcE7Xn1ej1ZvHWkqD12RmlO/Hwa5dq1a1hcXEQikcDLL7+MQqEAt9uN+fl5pFIpfO1rX4PH48FHPvIRPP7445Jr2t7exmAwwPz8PLxeL27evCn66vf7cfHiRSQSCXg8HjSbTcTjccRiMaRSKTQaDSwtLYmjMBwO4fF4EAwG4Xa7sbu7C7fbDa/XK3ktLgbEmD0ej3j2AMT+1Go1sWfnzp3D6dOnsbi4KNcldr6wsACfzydzeZIycaviDBu1h6M96U6nc4un5VQkjUU7Q1F+Rv/m+zyXDqX1+1pJ+frtcEgaciZKNKTCa+nPz0Jg2LxN7XFq40c8kfOCISbnCwAJW+nhBoNBgcSINerzOuEIpxwFywC3N9p6Pmkvmo4IIzPtEEw7BEa5dOkS2u02UqkU1tbWkEwmkc1m5Zk99thjCAQCOH36NHw+H9rtNhqNBlqtFuLxuDh5tBOhUAiZTAbLy8sCOw4GAyQSCTHYjHw5ToPBAMDYk6azRQ9Z2wZ+jk4EvW3aIADiTI5GI5w5cwbZbBZ+vx/7+/sYDAbiJLrdbqRSKTl+knIsxlp7slpB+IBGo5EMgg5rtTHlqnc7JdCf1Q+FSsdrOZN/fF3jl7y+VlR+lobC5/NJOMbB1UaCSkwvbCZjoUet8WCNC9I7pjI44Q9GKh6PB8lkEr1eT5JI+nzasHM8gFtx7dst+kcdqxcDXovXpQd9O4di2o32tWvX0O/3sba2hve+9704deoUEomERCZPP/20QBStVgvlchm1Wg2dTgfhcBidTgeNRkM81EgkglOnTiGdTqPdbqPdbsPtdiORSCAWiyEcDktSkL+phz6fTzByHflzrnFsmYwkHKLtlTbsCwsLiEQi6Pf72N7elu/MhYhJz0nLscAg2nvSN80H5HK5bJgQ32NCkQpOeMHr9d4SDju9IW3UnQqjvSytYPo1ht3as9MrsN/vx2g0QrPZRKPRkNWTCwoN+MyzBpLJJKLRqGDV9Xpd3tNJRM2+4PhTKYghRiIR+P1+hMNhrKysyLMulUrY2tpCq9VCt9uVxZ2LpTM5SXF63/xhnoWhrvaygUPPigrNRZvGh/PGueBPq8RiMTQaDWxubiISieD555+H1+tFr9fDwsICGo0G2u02KpUKKpUK8vk8ms2mQAi0DzTwmUwGiUQC9XpdIvNcLodz584J7BmNRtFqtcTLpZccDAbh9/sFBzfGIJFIIBwOS1Kx0+nInONiQXvV7XZlPPv9PkqlksCelUoFnU5HaIV6IW+1WhN9phM31lpRqDz8olRUhiihUEiMJL0orbBUZCq9Dk+dySVnolF7zhoP19Q83p82ztpY8z6YmAwGgzbIhMrNzzkhkmkVsms0jssxqVar6PV68qPhLuBwjtCTjsfjwsjJZrMyZpFIBOFwWJS9VCrJ3GB2XkdN2kt2wiccS2dEpw0wz8HkFY/VToXTy55m0Y6UMQaNRgORSAQulwtXrlxBMBhEt9tFqVSSPEYgEEAsFhNvmos48x60DzTYGnZ0u92Ix+MoFototVpotVpibIkns8bC7XYjm82KoaajBUAcslarJfoNQBaIcrmMnZ0dhMNhhMNhJBIJvPnmm1hYWEA0GkWpVEImk7klGT0JmbixHgwGt8AXNLoaAxwOh1KMogtdgEPv3HajDioOz8PfGvvU8Ium/mgcVXtwWnn18U6eJrFrMk70ouE0/NMsTiYGX9P0TUILzoiIY+Dz+STEJaskHA5LpEbebTQaRTQalYQTx8s5n/Tc0bAI54HOR2jjzXtyGnnnfCJWOmmc8qQKC0QikQiy2Sza7bZADRxDjhUTe3SICoWCsMOGw6Et8a8dwHA4LHqv4TJjjBzPiFjnFUajkSQpgcP5oJ2MwWCAVqslc8Hv90uCXCe+I5GIeNKE4vQcnqQci7HWCUUmYvx+v81YExsKBAJiADW8Adg5q0wMaOhEF1fwWnzg/E3YQ3u+fNCcIM5r8XPEvzqdjixCpKKxWg241SN3LjTTJjoZw/HRUYiGKI5KzjGDn8lkEIlEhE/PZ08MM5VKodPpoNlsIhqNIpfLiafthLk0PKeTh4A9IUqD64zQeH96bDlnKDOP+lC63S7m5uawtLSEs2fPYnt7WxJ80WgU3W5XPGIyOEipKxaLQpFrt9uIRqOi6x6PR5ynVCoFy7KkeG00GklVcr/fR7PZRCQSgdvtRrfbFQjNsixkMhmBNziGdB6NMSgWi2g0GnLP8Xgc3W5XzsP7CYVCkhCt1WrIZrMyxzQ1eBJyLBwzeqAulwvBYFAMJzErDoSTAqUNHo2sXv00tqmxLRoEXhuwe3d8cDyfTm5ygvAYjX9rcjsxLd5vv99Hp9NBLBZDKBSCZVlSRj3tQngAgOQnOp2OYIs0ikwE8XNUxnA4LJzZUChka0nA+aSN/nA4xNzcHBqNBiqVCvb29nD9+nXU63UbTAHYC3P4oyM5UrmAo42zdiacSWr93rTL0tISnnzySczNzeHSpUuo1WqIxWKSqO92uygUCnjrrbekVJwwI73Ufr8vRpz6XK1WJXIiXMK2EvV63VatXKvVBPvm+HN86EDQdrDAZTQaoVqtyty0rDEfW+fNCoUC4vE4gEO+P+fhwsKCnDMcnsSm64cycWMdCATk4TrhDcuybJVBxHWc3q7mShN+6HQ6tlJgKpETn9ZYo76uvobGIp1hq1NBj+Lx0nPXniI/64RqplG4kHL8AMgiTTqWxiU17s9kLgsRWLjg8/kQCoVsiqsXelabxmIxZP7/7Z1Zb1tpcv4faqO4k6I2S7K8ydMdzzQmGCRAMAiCABkgSG5ym4v5APkG+UC5y12uEgQzPckggyQznV7t7vYi2dZKivsmiqRI5kL/X7HOsToXiYR/G1QBAmyJPDw87/vW8tRTVcvLSiQSKpVK5mnzmYTEng/tIQ0PoXk2gI8AkHACOxwVTrPs7Oxobm5OzWZTjUZD+XzeznGn01GhUFCpVFKpVFK73dZ4PFY8Hlc2m9XGxoby+bztmfH4sn8P7Qk6nY7phVQqZb1BiNaBKHu9nlUUzs7OGhRDtB+G4/CcW62W+v2+YrGY0um0stmsFhcXrQQ+Ho8bpo2hkWRwiU90X6dcu7L2ijpMqQP+gHrl6VzhQhWPAYdDWv/aMBYVDnH5bB/uhEPyq2ALfwD9ZyHhYp2wAphmoewXeAr4yvda8LxoXsMBXFxcNA8aRQ0EclU+hPcTJeXzeUWjUWWzWVUqFR0fH6tararb7RpdkPeHE8zee/ZeGJ/l4Zsw7HHrVU9kZWXFImiU6MXFhXGpi8WiGo2GpCApgSgKpYsily6hFSJp1juVSimRSCgWi1kfEbxiyAm+UC0Mu6FPUObn5+eBxDe4M4oYBsn5+bl53X4/AH14iu91yY30BuGL0mTJN0E5Pz9Xt9tVp9PR4uKiGo2GYU1h5YeXDubNA8BDw4sOsz38gvgGUJ7Ly++8B0gyw8Mnc3NzAQ4v94CSRwHBzfQLPK1CIpDnORqNTFGStMVQe0N7cXFhijkej5sHRKIqXMbPXuGw4eWMx2OtrKxoc3NT9XpdBwcHOjg4UKvVUqvVMjySMNdHbFzXG3dvzMPeN3uK990q7EuJRqPqdru6uLhQMplUv99Xs9lUp9NROp1WuVzWwsKCdnZ2LJKemZlRpVIxo8uegIEDFBmPxxWNRrW+vm6l57FYTPV6XQsLC2q32zo6OjL2EPuIDpxAc9KloWi1Wkomkzo7O7PKxnw+r1arpXK5rGazqZ2dHYvcZmZm1G63dX5+bnxr35NoMBio2+2q2Wxe6zO9dmWNdfHgOg8dSk2z2bRmKGdnZ1a5BvaUyWQs08uP57Z65Ql+zKGheZNX1hwusso+c+yLXVAiPiwG615YWFChUAgkKVikXq9nobk3CtMqYUiBZ450Op0AU8cruVQqpUwmY1glyjoej0uaJC+pNmNtPBOA12WzWcXjcaVSKd25c0fNZlP1el2vX79WoVAwA0Kyi3X1njXemfQuR1sK9rq5VdYT4TkMBgPrqCddKrNarWaQYaVSCURSg8HAsOvZ2Vmtr69b9JNIJJTJZMzZonoxEokYcwPvmHtgbWKxmKRJMnlhYSFQNFMqlVQulzUej/Xw4UPVajXt7e1pd3dX0mVF5urqqj744AOl02mDY05PT5XP543PfXZ2puFwqHK5rN3dXf3VX/3VtT3TG1HWnl/pvZ9+v29lpfV6XZIsAQTgD8YFaZ3w1z94PF4PZ/hD5MMPlL1PQHpDAm7uGQpSsJ0m3yUWi5kn7RU6meEwVj6t4kNJDz0RovrK1TAFkoQt6xtOQgOh+MgKZR2GzPw1ZmdnjRebz+etX3qz2dTu7q4xfrg29+jL2vkeyO1af7eQgBuPxzo8PNT6+rpVLA4GA52dnZmxXFxcNKdtPB5raWlJuVzOimC+q781BAAfcfvydIgFwGfn5+e2zsViMYAvFwoF26MY9G63q3w+r9XVVQ2HQ+NWN5tN27/AInj+JDlrtZo5pNclN8IGQYmFlTUMirOzM2uKxOHodDqSpHq9bgMF0um0JQgkmffEAuB5eUYHHl24JNzfg6+khDTP9T1HF5iD77O4uGgevv+enrst3R5i1tJz1jGs3ov1rAy8WBggns7pIQeev6+O5Xn7NfUUPPYHFKx79+7p/PzcIrxOp6Nms2mGGGPDjzShIYaTz1553yYYJ8JgjvPzczUaDXv+8/PzKpVKarVaqtfrarfbWltb0/HxsRqNhubm5rS6uqrt7W2DK5aWlpRKpQxSI9ri+qwN+TBJAWVNwrvZbFrF68nJidbX1w1+rdfrWltbM2VdKpWUTqe1vb2tnZ0dW1fey/7K5XIWlfd6PavMrFarqlar1/pMr11ZQ8nyYSQPczweG/g/MzOjeDxu8AicSvClg4ODQDEEDyaTyVixhGeL8DA5TN5D9j07+Bz65zIAgfvlYWOFUcLQfAixfOc9fzDDOOY0SrPZtKIVKdiUCwoUXRdhaRAtYaQXFxcDzKHZ2VnL6pNoGo/HqtfrFr6urq7aIZYmCUScAdaMsJlQ9v79+3aI8dDYR/RVbjQaqlarhruHI7AwBDLtyjrcUAm2RyKR0NHRkXXhe/jwobU5xQk6ODhQsVhULBZTNpvV2dmZtUBlcAFl6NKEQdZqtczxAqKkSvqrr75So9FQp9MxhwF2R6PR0MbGhiQZpPmXf/mXyufzymQy5iyCu3/66aeSLplvGxsb+uyzzyTJytS5/+89Zu2rjQiFgUA8PS+RSGhzc9MODQlEzw5hQWCY0CdiOBxaOSkZWk+f6/f7dnjw4vDqK5WKedeJREJLS0uWDGu1WqrVaoF79TALi+VphbTOZJNMu6KWZNCE7xUsTQwneQBJ5s16hoanQ8LJ7fV6gSIYQkyKLNbW1iyk9YrZRz3sI4xDIpHQ9va2VaYRxhLm8h68pk6no/39fTvwMAJ8JOYhm2mWmZkZWytJRiyAx/7gwQNls1ltb2+rWq1apOWbMXnIot/v6+joSOfn5zo+PraS7t///d8PJP49Tff+/fuKRCKqVqsqFAoqFotqtVrqdruKRCI6OTmxdcpkMrpz544p53Q6bQ4aQxA6nY4p4WQyqXQ6rfF4rHQ6bR788fGx9eNOJpPX+kyvXVl7LJnEn7eu3uPwVBkOFzgnHpIvZkFxEvIQPqN8PUfah6Mcdrp1kQSYmZnRycmJef+UkvJ+8M6wwk8mkwEsnLLZWxjkUnxVKgk71tX/eDiEPcJ+wJvhmbdaLc3Ozlro7KlRKGHYB7BQiNJIcLJ/4OyjXPv9vpLJpLrdrmKxWKD/B/zb0eiyRHl+fl71el3NZtO6xPGZPlcy7XsgFouZ0iRyxfDSBjWdTlu5Nt4xugOotFarqdvtqtFomOIsl8uWI9rZ2QnAVxjKaDSqZDJpRpY9hKGdmZmxKTOsK8Uvo9FIrVYr4GSQKD04OFC329Xy8rKN+qOkHYN0cHAgSTZ/9rrk2pW1xxpRZJ7VwQL2+30dHBxYFSKUGU/7w3OFvhPm6HK9brdrCx3GyiORiOHkVBhi8Vl4X2jDpAkUjjSh93iSOwfT9w/xDIdpFhKE3nihkM/Pzw3GAH/2UUqr1TIWhzTxvGlHCZNkbW1N4/HYYJT19XVVKhUz9BRKsJaekgnERRTnlW2n01E+n7e9S0JcujQKm5ubOj09VbVa1cnJier1ujkdZ2dnUw9/IJR5SzKqnSTDcynRPj8/VzweVzKZtAhpaWnJ9sLh4aFisZgKhYIxR8CfR6ORyuWyRT+SDDal9weetDQ5s8Cm6Bzgs4uLCzPCnOe5uTmrjiyVSnrz5o16vZ6SyaSy2axBdOiTwWCgg4MDzc7Oamtr61qf6Y0kGFFYiUQioMzwnnkoo9HIRvL86Ec/CvSi5SH1ej1bpHa7bYUn9I+AW4v3BFHdFzTA+8VqsuC8p9vtmtGARjg3dzmxnAKL2dlZHR4eWpYaiIfX4ondyoTOhudKJAN8xdqQQyA3QCKQXsYo+1arZV31EomEUqmUIpGIlpaWzAifnJwY9ZOyZtYUbi+QGkUUOA3e0/cUQWiapVLJ2ArZbFbZbFbdblc/+MEPVK1WrYdEuVxWu91Wo9FQpVL5//b8vw9CleHCwoIymYxVMOJY1Wo1M9CcSSCt1dVVYwXl83n95Cc/MSijUqlYhLy0tKR6vW4eezwet+k0Dx8+VCQSMa8ZGLXVatl6vX37Vs+fP9cnn3yiu3fvWu7j/v37Gg6HOj09Va1W0/r6uuXhYrGY/uRP/sS85tnZywk06BNYMMViUf/+7/9+rc/0RmAQaVIlKAX5jj4xk0gkLLNLYqnRaKher1tyKhaLaW1tTdJlTT58SgwAPSd8QYVv3IR1pbk52NPc3Jy2t7eVTqcNGmk2m4GxQIPBQIeHh5JkzWc8N5vPkSYJDV8RN63iGR6SAkUPdFXz8BGGnNAZDzWXy72DCUsyfq4kwxdJVPd6PcMUwcfJabAnPcTFmnF9IA9ei4Hh+0ATTCaTGo1GWl5etvC3UqmoXC6rVqvZoZ1WaTQa1hN6a2vLlBnFKb6ftTfsHvpcWFjQ1taW7Re84ZmZGYuWqtVqoPBlbW3N8Gyg00gkYjAo9OGLiwttbm6q1+vp008/NXhtfX1df/qnf6rhcKhf/OIXevr0qSqVihVZPX78WEtLS2b8QRBoh1AoFCxiJCK7LrmR4QPSu5hd+MCx4cEI6TmLhxOLxdTv962PMZgTB8dXvpHF9wU5eHIUr4CVeUx8dXVVyWTSssTz8/PK5XJaXV1VJpNRuVzW0dFRAKqRgk2iPPbKZ057MyfyAShFDh8KkteAO/O8otGo+v2+2u22+v2+UqlUIMGLx07I2+12rfk/GCeRlR9mwP+5B2myTz29kH0Fh9vzuMEl2V8+J4LTQLjPfp5mYeAEz4PhtTTpInqiNzVUzKWlJaNTDgYDnZ6eqtfrmbe8t7cXOOM4dolEwii/mUzGoihJpk+oimX/bG5uqtvtam5uTuVy2VoUPHz4UKenp9ZHPxKJKJfL6c6dO9re3raiLai+OInz8/M6PDw0BhM0wuuSG1HWPswFE6ToxTdxolIRzAfrCs82Ho9bp7Vut2sPpd/v686dO4aFYikpHfYNorBueG2np6cWLi8vLxtNEIbKaDRSIpHQ+vq6BoOBnjx5YiFYrVYLhEPgpZKsLN1XtE2roKyhTjG1g30A7OBHL6FImYYdjUaVSqWUz+dNsS4sLJhhzufzVjJer9cDcNnm5qZxc8FDPY7uvX4MrlfafjK1Z4VIsmpKD7XhrfkOb9PecqDZbJoCrVarOj8/Vy6X0/LyslKplEajkY6OjnRycqJGo6FoNKrNzU397Gc/01dffaXnz5/bmdvZ2dFHH32kn//85/rbv/1bc9gSiYTRcKl2ZK1J/nEmOd8UrMTjcf3whz/UxsaG+v2+Pv74Y/3kJz/Rj3/8Y7Xbbf3TP/2TTk5O9PDhQ3300Ud68uSJlpaWtLi4qKWlJYPLarWaNjY2NBwO1Wg09B//8R9WFg9Wfl1yI/2sEV/lR7jAwVpaWgp0T8MS4tHgMY1GI1WrVQ0GA4NKUPYkK32fDj6LEBaFgQJdX1+3zyyVSmYRc7mcBoOBJSb4G9S+8XisUqlk4e7W1lagepHDjNc+zcIawNwhKRiJRCxxS2jqOc3AJCh2vCoMN/mMwWCgcrlszzoajarVaimfz2tlZUUbGxsGu4zHY6NyQickBxHm5rOHSFRz2GdnZ01Jg6sPh0PDLfHi5+bmVKlUjCY2zVIqlez57+7uanNzU7FYTKlUSvfv39edO3dUKpX0+vVr/epXv7Jzf3BwoO3tbTvXCwsL+vM//3PjY5MvSCQSevDggT7//HNLGPrKVx/5kuNaW1tTt9tVr9fTxx9/rKOjIy0vL+v3fu/39OGHH9pUomKxqI8++kgffvihZmZmtL29bbDa3Nyclbyj61KplJ49e6bPPvtMxWLR9vp1O203glnjPUkKUOz8KCcsYBgu8dQ7/s314OPCn/af6QtveIgofZrKUNEk6Z2F5LWecuS9Lt+XADYJ3hiJKhbJlzxPo3gjxoEl8mF4AMyQ8F4BzgK79jRJ328iDDVxWGk+jxdPqOphjfAh8obfJ4wxwn7aCElq7pEqularpePjY2v8c90d1943oVkb6+h79IA50w718ePHlnuSLiOzfD5vEOny8rIk6fT0VLlczqKm1dVVraysvJMPk4Lcfvq7wPyAiQaZQJLlqch5oZvm5+e1vLwcuA7OGOt/fHysFy9e6MWLF2Y4wjTl65AbmW7uOdWef9ztdi2sjMViSiQS77TNlIKDTP3veE2YvxuuYMQ78hg2r8Fr4pq+lBkPSZLxt/kckqDwgOv1utLpdCAUvlXWEwkbQiAPnrlX1lKw57hPCvmSbl9wRDESa+0PGt43lC68IDjzPtqDnsXB4t+Ez3wXjAPJZWnS7a/b7apcLuvg4EDlctkcg2kWMGmeJR4tThCRTj6f1w9+8APV63XDf1HW29vbhkcPBgM1Gg3rVZ5Op7W0tKSNjQ212207r+E6C/SDdLl/8vm85ufn1W63VSwWzTkgSck4OWnSRygej5vhhn3Gnpibm9Pbt2+1u7ur/f19a0B3E+ywG6HuUY3os6GDwUCvXr0y7JnuatLlwSuVSnbAgDB44BwUJi+QROIwnZ+fm8fGpGGSS7A08vm8crmcTk9PA1VzeH4sKoc1DKUsLCxoc3PTqFq7u7v64IMPLDTGSt9Wr8mgpGg0alnzUqkUgDX8sFLGK4EJsg5g04uLi9bLIx6PGzwBZtntdo0RQlUhhpyISZoYcj6P9WLtUNx+7X2Blx+YQK6k0+no6OhIb9++NfwV6GSahWIWIpXT01Mlk0nlcjnt7Oyo3+/bGX706JElFUkyAk965ZtIJHTv3r3AEIr79+9by1sqDXG8yCdxnil3z2QyWllZUb1etwpq77Shn3ytAOfa/wC5vHr1yhS/j8yuW26kKAYLymFpNBoql8s6PT21cU1wn73364tgfH8HPGOvCDkMHBoUr78uGKlvGrS8vGxWnmShJKNoAdOgxD2LIZ/P279PTk50enqqbDZrhTQs7LSzQXyOAm4sRQbZbNbYEzA+/uAP/kDZbFbr6+taWFjQ0dGR9Z9mSgeTpb3HBH+bpCBUMY9d+vDVV1JKk34e3nDjyeNx+UOHkeB9CwsL2t3dteEGiURCnU7HDNA0C43aaA9Rr9ft+d27d8/IBclkUp1OR9ls1hhbXmnS1Y6cAT2pyUXE43F75kS1Hib1+RCcQa6XSCSM0oeSl2RVlz6y4hoYd74LCW9J7yjq7/0MRg8x0KGqUqmoVqtJ0juwBEoYqh5hZvhQeY+H64dDDf7uvSpP+5ImU0x4sB6y8Qf2KmqXb9M4Pz9vVWsk0mg2hKGZVmFdKHB5/vy5Zmdnlcvl9OjRowBcdPfuXa2trVlvYjznRqNh7TPBoX1LTMTDGOwFru1/PMea//t1liZNuXitp/SxNzymTsIcTNbzs2/Cs3qfBFgQb7Pf76tarWpubk6FQsF6WOPBJpPJQJTFWfbKmpbJRDnUUYQdOS9+/TxtEzIDUbOP5NFD0ruUY/auNGn77Bu++eT6dcOh165VPF2vUqnok08+Ua1W03A41M7OTuCQsLEB9M/OzgzY9xxcv3g8AD8o11tUvGEW0DcJ4nNhJFBIIU0W1S84VhLlgNe9sLBgm4umMH6gLrSeaRWvrMfjsZ4/f26JWz/9Jx6P68mTJ8pmswGsslarmacF2wNIpNFoWNsAX3jFPvIH0h8sr6C91+09Me7bh8Rcm8MoTTji9XrdhqtSbDU7O2vFEtMs9PGA5irJWjbs7u4qn89bcYlfKw9Fcpb9kGwS199ldH0y2lM1wcMRlCrXR2/BNINjzbU8yQHBGYETDm4Oxv29x6wTiYT+8z//U0+fPtUXX3xhB29lZcVew6HAElHN5hfNwxdhhoincGG5CVv4HVTA0WhkwzR9Uxl/2DlYnkXCwngj4D0zQnM+72c/+5lOT09tpNA0SyqVss2azWb113/914E5e8Ph5Ri0u3fvKp/Pm/JLpVKW+CkWi9rc3LQeEyQkMYY0UcJLIvIaDoeGYxOCU3nqG9f75DUK3E+wRoiSyJOAeddqNcMqgWBqtZpRvKY9yexZYCSKR6PLBkm/+c1vrFNdJpNRKpV6JwL2feq9YvXRL7qB1hPkFtAJ3stlbVHKsHw4+0RvUhAeQzdhJDwRAocUiIcSeelSxxHFX5dcu7Le29vTb37zGz179syUGhbHN47nIXDIvAJHKXrsh0PkWR8++ePDYK9g+T/WGdgCLBwPCuvpPXHPUAiHtWwiWCIffPCBHjx4oMPDQ33xxRfX/VjfK/n8889tKvTW1pbu3LljBlaSlQLTOAlve3Z21iaSd7tdq2CE7QFNjiZQYNTsGd+ylIZR4XwI6+ijOs+Z9fsPiMzL+fm5edSj0UjNZtMqa2dmZiypPu3KmkI2T9mULs9tvV7X06dPFYlElMlk9Pjx40BlqI+8vRApoxtQnBhi/8x9hOQjLM4tOsi3H/CedBhexdEg8YzTeHp6arqIvYQh+d4r688//1xff/219vf3df/+fbthDqsvy+ZhoRDDDxbhd1hS72mzgEgYK/JhMAuFdfYHls/wWLh//VW/o/HP2tqa1tbWzAgcHx9f92N9r6Rer2t+ft48Jtg5eDVUphIJ4RXxXhJGVLcS6sbjcasK802YOGzgpFwXuuD8/LwdbF7jvXH2BuKNvIc/RqPLNrqNRsMakflkqc+zTLvE43Grq0ChIufn53r79q3S6bSePHnyTq4pjEH7548eYP05+8AnvC5M4ZP0znn2EKsnNPjXe/EeN9076VsSzo/wPa5Trl1Z/93f/Z2N5wnPLDw7O7O2gt5rlSZhpt/s4QdHBz8OhqQAHxYPzL/PJ4N8MxhJZpF9+BVW2L7fB9GAb++5tramH/3oR3bwaQQzzfLo0SMrNSbq4XBEIhGbHJ3L5QKVhr1eT4VCQa1WS3Nzc6bs2fjr6+uWhJqdnbUG9YTZ3tgDV3hKHkJE5IswPO0SJgdJZRgAnU5HlUrF4C5p0uMmGo2q0WhodXX1NsEo2bRwD1nyTIbDoQqFgpaXlw3XhgWGEWc9fF0F3jR7gFam7DHvvXsePd6w7/qJ/vH7K2wc+H/YK0cHtNtta/MsybBuTw29Trl2ZV0qlcxaNZtNo9pQPAAnGsDfQw48VP8QJdlC0h2vUqlYiIwV9lSbMIeajK1vFC9NFL0vxPCJJjwxjATd1cCqV1dX1ev19Nlnn+njjz82jncul9Pf/M3fXPejfW9kZWXF1oUudqzncDg0rqunYeFh04UtmUyq3+9bO9OFhQWdn59bwcLs7GU/mVqtZp44jZR4vX//aDSZVO9zJd4zw4v3TZ9Q+o1GQycnJzo+PtbBwYH1ifHDU/nsm+LZvk8CdQ9l6SsYeTb1el0vX77Uj3/8Y2MC+TySV3je2A8GlwN3mZWIosa4kqD0kTPnGoeNYipyFcBYXnHznrDDJl3yyA8PD7W7u2tQ3Gg0CsyWvW65dmXtWQDh4gMsYpjJgfhqNkkB5UubU3BIn53FGuNB+3JhnxQYj8dqNBq2IAw0wIMmfPaLg6L2HjWbIxaL6fDwUIVCQfV63ULuZDI51coaQQlLk4bvHJJoNKpcLmcMCpSkb9IThitQ/Pw+lUqZISfcBmqhHa5PXsdiMcvye7aRb8rkIyo8vrOzM5XL5UAzLylI/4xEIkqn0+8wiKZVYEhcRX3zeQUiKdaGylTv3XrnyyebK5WKdcvEq/Ywqo+0MLx43Z6YgAHxOHkYdkWf4FlXq1UVi0W1221rEMV+8ont65Qb6brHl/NcSJQ1OBZfztPxONjSpNsZ16lUKmZF7969awdOClY2DgaDwFgmH/5IsvFQJLX83/H8whaVe/UhF17A/v6+Pvnkk4CXNu3JJU+BwwhKChhDmmexPzhgYNIkjHyrUz9AV1KggrHf7we8eJ9UZM2gVQ0GA2OFXFVMgaKGjlev13V6eqqTkxNT1lC8SIpFIhHbhySfplmYxCNN4AOf+MeZIzKiNwgFTNIkz+TJBZxRDKjfI2HYIays/fn27CT+dlWS8SqFTSMxmED5fN6ibXI06LrrlBtR1jyMZrOptbW1gPcMnED/3zDLQ5r0AcE6tdttHR8fm9fKw8ArxqP1UIh/UD6RIF1CNXjnS0tLgU3kFTW4NIqaqCASiWh1dVX7+/tG28lms2q32+8kU6ZRwmO0/POnnSi0PiCsfr+vN2/eqNVqWfFDMpl8Z5YjChgFjfJGucN/5jX0oKE0PJFIKBqNWptTFDafJU0mc9OPolQqqVQqqVAoqFAo2AxIPCgKdyqVil2PQbHTLp6lhXBGSQpWKhXrWf3o0SOdnZ0FqLfeCRqNRjZqrV6va2dnR+l02qqIve7wOsV7zZ57TURGoycivDBOjfKForm/v6+joyOjgXJdhuSiM65TbrRFKiXdHhuCboVS9OGRt2wofDyx1dVVO1zhLnueKy29W+7tkwnLy8vG0WX+om+Z6b1x30OEz6IvQTQa1d7enur1uubm5tRoNCTJBnVOs1DWC9wEpQ4YhLWan59XLBbTzMyMms2mnj59qk6nY/8vlUq2fyTp9evX2tvbU6FQULlctrJyIBEMN2EyB7vT6SiVSpmTwFRqCpkkBbj5ftjE6emp3r59a1W4KBq+DxWXtCqQJonraRbOEOfORy/8vd/vq9Fo6Le//a22trZ0//59PXnyxM6epMAgbNpDlMtlHR8fG6QWjmRR2OgWzwTjfugl4yNmkpmeASTJ9hMVy5VKxdpNDIdDw6n5tx/IfZ1yI+Xm3rulOMEvFA/eJ/b4vTRRrvzfj+jxIYv32MLXD18TY5BIJGzIAH0BpElPYt4D/OGxazy4+fl5K6WHb+t7Gk+7+OKCMC1OmqwPLA2U6uvXr62ApdlsqlgsGtODqSEYWg47/SPozMZ1eQ0Og4ez/KRz1g5PKxqNWljOZOxms6l2u233hlPhQ2vfI4LvOM3ilaOnSHq4gcERL1++1OLiotbX122IrTSh4YZpfI1Gwxo/wewaDAaBHvPes8UJ8zqHMvZwrsNTe3kvf0dfVKtVNRoNi7D8eyExgIdfp9yIZvFULDql8TsWwhct8DcPVfDlPR3PQyZkeD0m5b1jb8G5p0jksjKS6SPMdPSbQQoqa5+QAI/y1p1x9isrK2ZA/KSRaZRwgsezQTCIfn2BG169emXebqVSMZhKkvGbZ2ZmrLnTaHRZNQaToNvt2kzPhYUFG6aL14viJt8QiURMwZKopkMgGCTrSZ4FjNQzS8IFOrd86yDRgDMZZlqA/dZqNW1tbRkOnc1mrQ+MJFN6nPFarWaT5cvlsp3/R48eBcgAfq8Bj0ERZoAvZ5b1lYK1GQiedbfbVaFQsBmTvM9zvz2Md51yI541GxV4gLHt0rtsESAIX7WIRSZhw8PnIbbbbfNk/bgl3zgcy41BwJsie5xOpw1nJDHAfXvYw2ei6fzVaDT0L//yL9bSkdAZLHbaD6qPenxixlMleUbz8/N68+aN9vb2VKlU7JC2Wi3V63XV63Ulk0ml02mtr68by2A4HBr7gkOVTCYDjAE84lqtZtejjSajmYDmaBoFxZS5nHhIZPzr9bpFeoTxQGlAHzQemmbxLBscGpwyICpeF4/H9ebNG3OK/uiP/ki5XC6gNL1DRa7g6dOnajQaBmHU63X9xV/8hR48eBCI5tvttp4/f65//ud/1ocffqg//MM/VCwWC9RXwJv2jCMfHeBV1+t1PX/+XEdHRyqXy9Yx0LeiyGQyto+uU27EsyYcnZubs2RAu902eo40aTcJ1OBDHuhQXvHzHo8/+ynI3qu2L/f/euPC8YSfiUfnQ2dJ5mn5a/hNkk6nVa/X9fbtW71+/doYBdM+by8s4ZyCZ9z4cl0YFaVSSUdHR+r1eoZhE1HRBIj+yLA34GEDR4SLXKRJkcLi4qK63a4Z70qlYslOPoO/Y9Dh0wOFYLDxysfjcaBFKwcc52PaGUGeAseoNhwunhV7hGIoqL6rq6s2R5MmX5zD0Wikx48fm0Gt1+tmlH/961/r/PxcOzs7evz4se7cuRPo6vfHf/zH2tzc1Pr6eqCwbnZ21u4Riih6QpLpqG63q1qtppcvX6parVqewxdUeaMdzp39X+VGedZAFd1uV81mU7lcLsC79D8e5/QUHawbC+UTkL7NqreEeHB4PHT7gs3h8UaoXHx2OHzj9xiHcrms/f19NZtNZbPZAIzjqV/TLOGMvIeU8Iow1LFYzKoCPb2LtWHifavVUrlc1ng8tqgKb83DaBhlPsPvF5Q1SWHuiWsQBgOddDodtdttnZ+f2+f54imwSWniSBAlXjcT4H0TnhPnM+wE+bzUYDCwasTZ2Vnt7e2Zckf5eWeMAbWVSkXPnj0zz7dYLOrzzz9XrVYzPHlra0u5XE7ZbFbLy8tKp9Nm6MP3CJMHaMPvJ0lGSjg5ObG+2xRgsYdIet8EK+zGMGvp0iJR6FAsFrW+vm4K2B8iX1vvk4ZecftDwnt8pRIK0nNwgSjwjkg6eczMz9+TgsaG74CCuLi40MuXL/Xtt98GNhDiN980y9nZWYBj7ylSVK1SINVoNHR8fKxCoRBo1j4zM6NUKqVsNmsc5oODg0DbAGh4c3NzNkKKvYE33Ov1VK1WVavV7BDC3GHfsdYoFehZNGxifBsFO0QGePg+Ec36X3cI/L6J7zxIwzbOlafd+upGSSoWi/rd736nk5MTbW9vK5VKWTTOWc3n80qlUlpZWdGjR4/06tUr7e3t6cWLFzo7O9ObN2/U6/X06NEj5XI53b17164BYwjYEx3DvXDOPSwKhbDZbNrU9VgsZvCuV8p+7Nv3vpETStQXRFCxuL+/H8CuSPaQIKCvrJ/WQiY2TFIHW2LRye7yg4IoFAp2cD1Xl0Pl5/z5hfMRANO1/+Ef/kEvXrxQu902TAwGCJlgn6iYVsEQchgwbOHOe/Pz8yoWi9ZpD+823L1sdnZW8XhcP/3pT/XixQsLQdvttg1lZZ+gNOkzDAfeT1LHWyLii8VihmVLky6NiUQi0MkvGo0Gysu5X8JfHAPC6mkWvM3hcBjo9+xxYgyf1wcMs0WBMw6MCIjhxdIlLPnBBx/o7t27+ulPf2rUPvIbuVxOmUzGqlb9/pAUSAjDIGq1WraGfk339vb07Nkz/eu//qtmZ2ettSvcbs8G8oyh65Qb8aw9dc4rvUqlolarZd4QRSce3vDX4Pc+0eBx6TAORgiKVy0pMHnCW0vPTOA+wx491LDBYKD9/X29fPnSqHo+5OVevHc1zcJa+OIHj117CiaK00dXFEt4JsloNDJvajQaqVwuq1qtSgo2+WJtGN0mTdhJ7IFYLGZ0vXDHPM+xJ1T2mX4gDpQ0TgavY+2nXVmT/+E5hemcYb46fx+PxwY98Sz9iDBJgYiW3i/SZMAttQ4o6PF4bE6iZ5Kx/t1uV2/fvjVHgryJJMPU9/b29ObNG7Xbba2srAR0DDqA7+u7eF6n3KiyDi9CrVYzKt/i4qIVvcBvZXGANSgT9slHxIfM3ovDynnvzCcopMmGwfpxwD0nW5LRcNrttr755hsdHx+bkpYm3gNeevg60yx+rQh1UXSsDc/fwwooce+dAF3Nz8+blwxuiPj9IsmGoM7MzKjVatk1fbZemlSa8VmwAjw0ArQiTXpV4Clyb7zuqu8/rcL5vkq+S5nBHsF4S5cNn3q9npLJpCKRSKCqlQiXLp++rzUGezQamUJHD2CMz87OVK1W9ezZM0WjUaVSKW1ubpqjAK305cuXOjw8lCSj/vp7liakCWlihK5Trl1Z42H4zY4CGw6HKpfLFmJS/CDJehqjcHlIWNjV1VVT8r7E0+PNPCwSQtFoVPl8/h0WCdfn2p426HHxdDqtYrGoL7/8Ur/4xS8s7B2PJ5WMJCTw4MHkplkYKOCVraR3vGY8VEnvbGyeo+/7QMEB1/KJvnDFqU9Usu7ZbFZLS0vGwSaqa7fbRvvkgPr2BfV63Sh8RFZUKXrGQDweN1YBuPi0is9N+dYPeLMIUIlXePz0+30dHh7as85kMrpz54550LFYzBJ6nsfvk5KwNWh/AV+avumcebBx1o891el09Omnn+rLL79UvV7X3bt3A86opEA+DAbcTciNeNYeUvALIUnNZtPCFB8Oo4jBgGOxmDVI6Xa7Oj09tZCTQ8F0ZPi2PtTG8yEcwoNiqgeZ/mg0qnQ6HUg0+hD4d7/7nfb29iTpyuw035dD7stcp1VgAKEwWVMKSKRge4Gwl+VbAHjPGlyx3W5b3w9Py/RrmEqlAtROesrQbxgWEB4bhvjs7MwKnzD+JEWbzabN3ETJeCfD51qmvZKV83lxcaFyuRyo8pOClFiM5Hg8tjmLhULB8g2esz4YDLS+vq7l5WWtrKwEOnh2u12Dy3gNnP29vT2jh3LdtbU1bWxs6P79+9ra2lKn07G/93o9tVotFQoF/fKXv9TR0ZEpdZ8QJT/i+3b7rozXKTeyo3ySzkMhMzMzlvhptVpKJpNmJX0DHxQ4BwHvB0UIFxJFTpjjPWQ8LDx3FDC9j9kY9MH11hyF/+bNG+3v76tSqdj3wuMChglXOvI9p1n8swAC8fQ6xOcHwnLV7/HS8GI8dOKni0SjUUs4U9GI5+zhF6AZ9oMvR/eGmD0M/urvxxttvuMtfTO4fj7X5KFI6WrMejQaWQ8OKhqJcijzPjg4UCaTCRhVeO/0lL97967lxsiXQcHE4YtGo1YvgbKlB8jJyYlevnypN2/eWNuMXq8XcAquaknB9/ne86ylCSYc9jDBLDudjhqNRqDzWbvdDkAiHDq8sWQyaZjkaDRSo9GwTDLDCGKxmDWax8vFw2YhhsOhGQV40n7yS7fbtbDp3/7t33R0dGSlxCQvSJD6bnwe7552r4qNygH01af+99LV8Ie/hqeAAXEBX/kmYZ5JQpQGZOZ7WPOZQCORSMQwau9UILBY2GvhtfXfBWUNLXGahcScx6xR4JwVny/iOXJeeX6lUsn49nRp/Oabb1QsFjUYDKzBm38Pyntra0tra2taWlrS8vKyksmklpeXtby8bMqfaVZeP5yfn1v7g//6r/9SqVQy7//s7CxQ2OUTqDDHfO+h65Qb7Q3iqTmez9rtdnV8fKxsNmsYYCwWU6PRsPcRDnmuIw9lYWFBa2trBn3A6/XWFWzU93FAyfJ/3/2P96ZSKe3t7enp06d68eKFKYOLiwvDtUhwtNtt24xh/vg0S7lctudENMUz92PXCCv9ZBdJ73i/kmy6kO/14avH8KhYd5wF389DCvaaxrCCUfv30keEykbWlP3gmw1xXfrMQOOcZmm32xYVewIAkVA4CvVRFgrw4uJCr1+/ViwWUy6X07179/Rnf/ZnKpfLOjo60qeffqqNjQ0tLS1Z0jifz5tTSB8gxq6NRiPVajWVSiW9fPlS29vb5i2jQ4jAX716pS+//FJffPFFwIEgGc13gFVEBBeJRMwAvBfK2h/GsALzIWWtVjOlWa/XrTTcczK9xUI5++svLCyYJx5+D68ldPFlodIkWw30sbi4qGKxqOfPn+vp06eB7wSrBE/PDwD2CdWboOy8b+LpWZ6i5WED1iKVSpknTK8V6dKLpm85hp7sPc2erqKD+USx/3ygMaAO5CqaKZ8Bx9rT9MLKmmtgGDAK0w6F0QvIR56cS38+wjqC5yxNaHy0JM3lcqYcaSXBMIhut6sHDx7YOSWXxXpQ0l4oFCwSJ3IPQxm9Xk9Pnz419pdPgnoGGPfvsWl+73XCdcm1K2uvSMP4nv/3cDhUrVYzzisWDgXqebjeq7lKGaLI/Wd7fNtzsNkwHFpfWjo3N6fXr1/r7du3gdAHrNofUt/XJAz3TPtB9UnDsCLEm+Vvi4uLBktJCih3DoVn2WBceQ3X5VD6zwnf01X3CbMjzBLy/Guu5ymFUpDp5KNI/z2mVVCqMHOkq6eFe2E9OWMY6EajoWKxaMwPGrwtLi6qXC7btKBMJhNg/zC4hLN6fHysRqOhfr+vzc1NU7yedksnwNevX1uXPogSfr/5WgtPI/bVmde9B25EWbMIPtzxN8/BOz09tYOSzWbVbDZNIWL1WHC656HY+RvJAE/V8ooV9ojHl/3IIQ7YaHTZf+KXv/ylGo2G4WCE8J6tMhqNjAvqFfVVzIZpFMJJf+g8E4Dng9ecTqeVTqf19u3bAAVLku0PvGLgiXC0tri4aFg21+CQwjZi/3iM8uLiwiI6yt+JmsDMfT91fzClSbMez1q5lUvWF04TlaPSdztxYWaV/z9Vo+12W/1+X7lcLjAEpNFoWDLyhz/8oXK5nFWl+ja58/Pz2tjY0MOHD7W9vR2gCSKtVku//e1vtb+/b4wmL56MAAxCtA0LCuUf3iv/V7lRz1oKtsv0EAYHslQqGd+RkT4LCwtaXl62gxeLxczqDQYDtdvtdxrIh7nUlIqGRzV5yytNsNBSqaR//Md/VKvVCiTBUNT0OqaSzeOY4e897V4VSppn5Q2p92KBJvCuw8mnwWBgXe7AGykpDjN/JBm7COXqPXjP9EHYgx5PZc+hnKmI9ROs/ft95SWlx76IZlrlKjgS8RgwwjP0Z8/nGuiCKElra2tKp9NWqwHfvVgs6uTkRIuLi0qn01Zuns1m9fDhQ929e9fwazBtaZJvev36tb744gv96le/CrS4CCe9cebQQeQogOm49+89dS9M2/M4cnjRpMsH0Ov1dHx8bBjl/Py89cDl/R6zlib191dRwghPwuGXT1zx0+v19PLlS+3u7urw8PAdqw6bgM8P95gIK2z/udMqUOKAn3yjfq+M2Q/ee/VK0itkoCu8YSiX3lPz+QPe4wut/Pu9B8+gAdZWmqwlJc4oX38A+Z3fM7ftBi7FJ+7DOZ1wBHrVOeL3rCNnio6J/I79lclkbI4mzBEYIIlEwpQ36xceelKv1/XNN9/oq6++MlYa9+D/TaTg54x6Nhk5Den6pwXdmGd9VcLNb2pPnxoOhzbElszq0tJSAO7wJeP+IOOxhQ2Db3GIR817/UaqVqt6+vSpnj9/HoA/uI5X1IR1/1MJ7U1Y1PdN8KYxXJ614cNODhzN+j3Dgk2P1xOGmzw2jJJn+AO4oa9swwumUZiHNgivPXyDEqCLGnvJ9y4PH+SrlMu0Ckr6KqjwKgnDWvzOK/vZ2dlAEjqRSJhBzWQyWlpa0urqqrLZrHK5nJaWlgIRkT/7eNWc55OTE3399df69ttvLWfmxesV32IAzNvDZkTg3/vhAzwQKTiDL5wN9haWcOfk5ESVSkVv3rzRYDBQJpOx0NKXHft/eyqeP5iEvGDR/n5arZYp7r//+7/X8fGxzs/PlUql1O12bTGonoIFQhGOr86T3sXdph235BmOx2Nje/hmRxwQ1p8EY7vdtrBWklWRYRxzuZw2NjaMly/Jql2XlpZsiKpvGBWGpzysRoFNuEw9FosFuNvcpzcmJJ08c4FrAMVMs4SxYJ5ZGFaQJooQ8X9HQXoIrVar6fz8XMlkUg8ePHhnQo9vFMfeGY/HgbqOaDRqZ7lYLOrjjz/WV199pWKxaFRd7o3oDv2VSqWsmIo5s+gMFDUjDa9Trl1Ze6qU/50PRxHvDaPEofT9+te/Vjab1c7OjtbW1syiYSXxwjkknhLmlQK0P5JNZ2dnSiQSKhQK+uSTT3R8fKzR6LIpPlQtlLOnaUUiEfPAvMGRFDA6ni42rUJeoVqt2nDRZDJpFEs/fJQBtdKkbzAeSbPZVL1eN8VOyXg8Hrf+L9LkcCeTSbVaLUtUS5M95j13jytjVLxjgWftmUQI2DT7EKPBZ3hoZJolnLwLP4//KREf1h3kJXwTsF6vpxcvXmhubk7379/X6uqqpEmlsm/FG4bTeE2j0dDR0ZG++OILffbZZ1bh6PMNHrLh381mMzDImb+zT+B5f1cU8b+VG1HWhKkeK0b8QoQ9U94DrW93d1fxeFyrq6sByoyHNbgOD9J78Vw7zEbAmn799dfq9/uBhBgPHYXNwSVBhaL2n+cX9FZZSycnJzadZ35+3qrQgLJ8q4BIJKJSqaSTkxNJk6KV4XCoarWqTqfzDm0K7zsMhzCglCIH1iJcdehhMO85Y+glBT7rKkhPmtA3cUZ824FpV9bId3nP3yVXOXQeXvLnju6LjGZLJBKBs+4xbSIj9Ee/37epT99++60lr6+6h3DkzPuJrtiDPq/hE9/XJTeCWYez5tJ3K2z+HYYVhsOhXr16pVQqpY8++uid9/tObj5E8li4NOE98jnRaFTHx8fa29vTt99++04ygEPre0oAq2BF+X5eWbMxfIJhWuXZs2fW9IjuhR5WIJEH7dIrOZ4108z9oaFVAa0saYPKmpHpJzq7KnHp8yd4WdJkGAI0Twph/JRtKcjhhwHAd2OPeGdhWgXHyztUeLeS3lHI/n0eRpWCkIqviJ6bm9PBwYHBWPfu3Qtciy56PmkNja/b7erw8FAvXrzQl19+aXRNj21fBdew18LeNNAbxvt/ihz+txKZ9k11K7dyK7fyPsh0x+u3ciu3civvidwq61u5lVu5lfdAbpX1rdzKrdzKeyC3yvpWbuVWbuU9kFtlfSu3ciu38h7IrbK+lVu5lVt5D+S/AVRzxTSydQ+XAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<Figure size 432x288 with 9 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Get a batch of images and labels from the test generator\n",
    "x_batch, y_batch = next(test_generator)\n",
    "\n",
    "# Make predictions on this batch\n",
    "predictions = model.predict(x_batch)\n",
    "predicted_classes = np.argmax(predictions, axis=1)\n",
    "true_classes = np.argmax(y_batch, axis=1)\n",
    "\n",
    "# Visualization\n",
    "fig, axes = plt.subplots(3, 3)\n",
    "axes = axes.flatten()\n",
    "for i, ax in enumerate(axes):\n",
    "    img = np.squeeze(x_batch[i])  # Remove channel dim if grayscale\n",
    "    ax.imshow(img, cmap='gray')\n",
    "    ax.set_title(f\"Pred: {predicted_classes[i]}, True: {true_classes[i]}\")\n",
    "    ax.axis('off')\n",
    "plt.tight_layout()\n",
    "plt.show()"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3 (clean)",
   "language": "python",
   "name": "python3_clean"
  },
  "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
}