examples/time_series/cases/7.ipynb

{
 "cells": [
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "467482e4",
   "metadata": {},
   "source": [
    "### Example 7\n",
    "\n",
    "Refer to [/examples/time_series/generate_data.ipynb](https://github.com/serhatsoyer/py4ML/blob/main/examples/time_series/generate_data.ipynb) for the artificial time-series dataset (used in this example) generation  \n",
    "\n",
    "|Case|LSTM Shuffled|Dense Shuffled|LSTM Not Shuffled|Dense Not Shuffled|\n",
    "|-|-|-|-|-|\n",
    "|Example Number|[2](https://github.com/serhatsoyer/py4ML/blob/main/examples/time_series/cases/2.ipynb)|[3](https://github.com/serhatsoyer/py4ML/blob/main/examples/time_series/cases/3.ipynb)|7|[17](https://github.com/serhatsoyer/py4ML/blob/main/examples/time_series/cases/17.ipynb)|\n",
    "|ID|0_1_0_3_128_1_0_1_0_0|0_1_0_3_128_1_1_1_0_0|0_1_0_3_128_1_0_0_0_0|0_1_0_3_128_1_1_0_0_0|\n",
    "|Total params|17,185|18,753|17,185|18,753|\n",
    "|Total time (min)|1.5|0.3|2.3|0.3|\n",
    "|Num of epochs|49|37|78|26|\n",
    "|Time per epoch (s)|1.79|0.51|1.78|0.60|\n",
    "|Train RMSE|0.53|0.63|0.57|0.79|\n",
    "|Test RMSE|0.58|0.88|0.64|0.98|\n",
    "\n",
    "- LSTM performs better than MLP when there is a relatively complex within window input/output relationship\n",
    "- Shuffling inputs improves the performance both for MLP (multi layer perceptron or dense) and LSTM\n",
    "- MLP is much faster than LSTM with similar size in terms of training times\n",
    "\n",
    "Previous example: [/examples/time_series/cases/6.ipynb](https://github.com/serhatsoyer/py4ML/blob/main/examples/time_series/cases/6.ipynb)  \n",
    "Next example: [/examples/time_series/cases/8.ipynb](https://github.com/serhatsoyer/py4ML/blob/main/examples/time_series/cases/8.ipynb)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "6a748aa8",
   "metadata": {},
   "outputs": [],
   "source": [
    "import sys\n",
    "sys.path.insert(0, '../') # To be able to reach 'functions.py'\n",
    "sys.path.insert(0, '../../../') # To be able to reach 'datasets' folder\n",
    "import functions as fn"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "f978cd32",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Do the settings in this cell and run all\n",
    "class Train_Sets:\n",
    "    vers = 0 # 0 or 1\n",
    "    within = 1 # 0 or 1\n",
    "    inter = 0 # 0 or 1\n",
    "    numOfSens = 3 # 1 or 2 or 3\n",
    "    batchSize = 128 # (2 ** 0) or ... or (2 ** 8)\n",
    "    useHalf = True # False or True\n",
    "    dense = False # False or True\n",
    "    shuffle = False # False or True\n",
    "    stateful = False # False or True\n",
    "    resetStates = False # False or True"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "2fb5b2fc",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "ID: 0_1_0_3_128_1_0_0_0_0\n"
     ]
    }
   ],
   "source": [
    "Train_Sets, Paths = fn.init(Train_Sets)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "e3350324",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "X_train shape:      (9216, 32, 3)\n",
      "y_train shape:      (9216,)\n",
      "X_val shape:        (3072, 32, 3)\n",
      "y_val shape:        (3072,)\n",
      "X_test shape:       (3072, 32, 3)\n",
      "y_test shape:       (3072,)\n",
      "Common data type:   <class 'numpy.float32'>\n"
     ]
    }
   ],
   "source": [
    "X_train, X_val, X_test, y_train, y_val, y_test = fn.split_data(Train_Sets)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "6ad3409e",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Metal device set to: Apple M2\n",
      "\n",
      "systemMemory: 8.00 GB\n",
      "maxCacheSize: 2.67 GB\n",
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "2022-12-11 22:56:43.687208: I tensorflow/core/common_runtime/pluggable_device/pluggable_device_factory.cc:306] Could not identify NUMA node of platform GPU ID 0, defaulting to 0. Your kernel may not have been built with NUMA support.\n",
      "2022-12-11 22:56:43.687479: I tensorflow/core/common_runtime/pluggable_device/pluggable_device_factory.cc:272] Created TensorFlow device (/job:localhost/replica:0/task:0/device:GPU:0 with 0 MB memory) -> physical PluggableDevice (device: 0, name: METAL, pci bus id: <undefined>)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/128\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "2022-12-11 22:56:44.268278: W tensorflow/core/platform/profile_utils/cpu_utils.cc:128] Failed to get CPU frequency: 0 Hz\n",
      "2022-12-11 22:56:45.207630: I tensorflow/core/grappler/optimizers/custom_graph_optimizer_registry.cc:114] Plugin optimizer for device_type GPU is enabled.\n",
      "2022-12-11 22:56:45.379098: I tensorflow/core/grappler/optimizers/custom_graph_optimizer_registry.cc:114] Plugin optimizer for device_type GPU is enabled.\n",
      "2022-12-11 22:56:46.341904: I tensorflow/core/grappler/optimizers/custom_graph_optimizer_registry.cc:114] Plugin optimizer for device_type GPU is enabled.\n",
      "2022-12-11 22:56:46.657234: I tensorflow/core/grappler/optimizers/custom_graph_optimizer_registry.cc:114] Plugin optimizer for device_type GPU is enabled.\n",
      "2022-12-11 22:56:47.269089: I tensorflow/core/grappler/optimizers/custom_graph_optimizer_registry.cc:114] Plugin optimizer for device_type GPU is enabled.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "71/72 [============================>.] - ETA: 0s - loss: 19.9798 - root_mean_squared_error: 4.4699 - mean_absolute_error: 3.4505"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "2022-12-11 22:56:49.323738: I tensorflow/core/grappler/optimizers/custom_graph_optimizer_registry.cc:114] Plugin optimizer for device_type GPU is enabled.\n",
      "2022-12-11 22:56:49.389424: I tensorflow/core/grappler/optimizers/custom_graph_optimizer_registry.cc:114] Plugin optimizer for device_type GPU is enabled.\n",
      "2022-12-11 22:56:49.458312: I tensorflow/core/grappler/optimizers/custom_graph_optimizer_registry.cc:114] Plugin optimizer for device_type GPU is enabled.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "72/72 [==============================] - 6s 33ms/step - loss: 19.7752 - root_mean_squared_error: 4.4469 - mean_absolute_error: 3.4287 - val_loss: 4.4203 - val_root_mean_squared_error: 2.1024 - val_mean_absolute_error: 1.6686\n",
      "Epoch 2/128\n",
      "72/72 [==============================] - 2s 24ms/step - loss: 3.7212 - root_mean_squared_error: 1.9290 - mean_absolute_error: 1.5272 - val_loss: 2.6961 - val_root_mean_squared_error: 1.6420 - val_mean_absolute_error: 1.2874\n",
      "Epoch 3/128\n",
      "72/72 [==============================] - 2s 25ms/step - loss: 2.3552 - root_mean_squared_error: 1.5347 - mean_absolute_error: 1.1981 - val_loss: 1.9251 - val_root_mean_squared_error: 1.3875 - val_mean_absolute_error: 1.0641\n",
      "Epoch 4/128\n",
      "72/72 [==============================] - 2s 24ms/step - loss: 1.7577 - root_mean_squared_error: 1.3258 - mean_absolute_error: 1.0275 - val_loss: 1.4154 - val_root_mean_squared_error: 1.1897 - val_mean_absolute_error: 0.9197\n",
      "Epoch 5/128\n",
      "72/72 [==============================] - 2s 24ms/step - loss: 1.2880 - root_mean_squared_error: 1.1349 - mean_absolute_error: 0.8893 - val_loss: 1.0387 - val_root_mean_squared_error: 1.0192 - val_mean_absolute_error: 0.7994\n",
      "Epoch 6/128\n",
      "72/72 [==============================] - 2s 23ms/step - loss: 1.0399 - root_mean_squared_error: 1.0198 - mean_absolute_error: 0.8057 - val_loss: 0.8859 - val_root_mean_squared_error: 0.9412 - val_mean_absolute_error: 0.7365\n",
      "Epoch 7/128\n",
      "72/72 [==============================] - 2s 23ms/step - loss: 0.9085 - root_mean_squared_error: 0.9531 - mean_absolute_error: 0.7546 - val_loss: 0.7965 - val_root_mean_squared_error: 0.8925 - val_mean_absolute_error: 0.6953\n",
      "Epoch 8/128\n",
      "72/72 [==============================] - 2s 26ms/step - loss: 0.8185 - root_mean_squared_error: 0.9047 - mean_absolute_error: 0.7176 - val_loss: 0.7363 - val_root_mean_squared_error: 0.8581 - val_mean_absolute_error: 0.6682\n",
      "Epoch 9/128\n",
      "72/72 [==============================] - 2s 25ms/step - loss: 0.7576 - root_mean_squared_error: 0.8704 - mean_absolute_error: 0.6906 - val_loss: 0.6920 - val_root_mean_squared_error: 0.8319 - val_mean_absolute_error: 0.6476\n",
      "Epoch 10/128\n",
      "72/72 [==============================] - 2s 25ms/step - loss: 0.7108 - root_mean_squared_error: 0.8431 - mean_absolute_error: 0.6683 - val_loss: 0.6590 - val_root_mean_squared_error: 0.8118 - val_mean_absolute_error: 0.6321\n",
      "Epoch 11/128\n",
      "72/72 [==============================] - 2s 25ms/step - loss: 0.6718 - root_mean_squared_error: 0.8197 - mean_absolute_error: 0.6492 - val_loss: 0.6404 - val_root_mean_squared_error: 0.8003 - val_mean_absolute_error: 0.6251\n",
      "Epoch 12/128\n",
      "72/72 [==============================] - 2s 24ms/step - loss: 0.6410 - root_mean_squared_error: 0.8006 - mean_absolute_error: 0.6336 - val_loss: 0.6261 - val_root_mean_squared_error: 0.7913 - val_mean_absolute_error: 0.6195\n",
      "Epoch 13/128\n",
      "72/72 [==============================] - 2s 24ms/step - loss: 0.6176 - root_mean_squared_error: 0.7859 - mean_absolute_error: 0.6219 - val_loss: 0.6154 - val_root_mean_squared_error: 0.7845 - val_mean_absolute_error: 0.6156\n",
      "Epoch 14/128\n",
      "72/72 [==============================] - 2s 26ms/step - loss: 0.5958 - root_mean_squared_error: 0.7719 - mean_absolute_error: 0.6107 - val_loss: 0.6011 - val_root_mean_squared_error: 0.7753 - val_mean_absolute_error: 0.6091\n",
      "Epoch 15/128\n",
      "72/72 [==============================] - 2s 26ms/step - loss: 0.5774 - root_mean_squared_error: 0.7598 - mean_absolute_error: 0.6009 - val_loss: 0.5933 - val_root_mean_squared_error: 0.7703 - val_mean_absolute_error: 0.6065\n",
      "Epoch 16/128\n",
      "72/72 [==============================] - 2s 24ms/step - loss: 0.5619 - root_mean_squared_error: 0.7496 - mean_absolute_error: 0.5924 - val_loss: 0.5817 - val_root_mean_squared_error: 0.7627 - val_mean_absolute_error: 0.6010\n",
      "Epoch 17/128\n",
      "72/72 [==============================] - 2s 24ms/step - loss: 0.5476 - root_mean_squared_error: 0.7400 - mean_absolute_error: 0.5841 - val_loss: 0.5751 - val_root_mean_squared_error: 0.7583 - val_mean_absolute_error: 0.5986\n",
      "Epoch 18/128\n",
      "72/72 [==============================] - 2s 24ms/step - loss: 0.5358 - root_mean_squared_error: 0.7320 - mean_absolute_error: 0.5778 - val_loss: 0.5651 - val_root_mean_squared_error: 0.7517 - val_mean_absolute_error: 0.5943\n",
      "Epoch 19/128\n",
      "72/72 [==============================] - 2s 24ms/step - loss: 0.5250 - root_mean_squared_error: 0.7246 - mean_absolute_error: 0.5718 - val_loss: 0.5537 - val_root_mean_squared_error: 0.7441 - val_mean_absolute_error: 0.5882\n",
      "Epoch 20/128\n",
      "72/72 [==============================] - 2s 24ms/step - loss: 0.5142 - root_mean_squared_error: 0.7171 - mean_absolute_error: 0.5654 - val_loss: 0.5447 - val_root_mean_squared_error: 0.7381 - val_mean_absolute_error: 0.5837\n",
      "Epoch 21/128\n",
      "72/72 [==============================] - 2s 24ms/step - loss: 0.5052 - root_mean_squared_error: 0.7108 - mean_absolute_error: 0.5601 - val_loss: 0.5381 - val_root_mean_squared_error: 0.7336 - val_mean_absolute_error: 0.5802\n",
      "Epoch 22/128\n",
      "72/72 [==============================] - 2s 25ms/step - loss: 0.4978 - root_mean_squared_error: 0.7056 - mean_absolute_error: 0.5557 - val_loss: 0.5324 - val_root_mean_squared_error: 0.7297 - val_mean_absolute_error: 0.5773\n",
      "Epoch 23/128\n",
      "72/72 [==============================] - 2s 26ms/step - loss: 0.4897 - root_mean_squared_error: 0.6998 - mean_absolute_error: 0.5511 - val_loss: 0.5241 - val_root_mean_squared_error: 0.7239 - val_mean_absolute_error: 0.5731\n",
      "Epoch 24/128\n",
      "72/72 [==============================] - 2s 23ms/step - loss: 0.4815 - root_mean_squared_error: 0.6939 - mean_absolute_error: 0.5463 - val_loss: 0.5175 - val_root_mean_squared_error: 0.7193 - val_mean_absolute_error: 0.5692\n",
      "Epoch 25/128\n",
      "72/72 [==============================] - 2s 23ms/step - loss: 0.4740 - root_mean_squared_error: 0.6885 - mean_absolute_error: 0.5420 - val_loss: 0.5204 - val_root_mean_squared_error: 0.7214 - val_mean_absolute_error: 0.5716\n",
      "Epoch 26/128\n",
      "72/72 [==============================] - 2s 23ms/step - loss: 0.4673 - root_mean_squared_error: 0.6836 - mean_absolute_error: 0.5382 - val_loss: 0.5188 - val_root_mean_squared_error: 0.7203 - val_mean_absolute_error: 0.5711\n",
      "Epoch 27/128\n",
      "72/72 [==============================] - 2s 23ms/step - loss: 0.4612 - root_mean_squared_error: 0.6791 - mean_absolute_error: 0.5346 - val_loss: 0.5131 - val_root_mean_squared_error: 0.7163 - val_mean_absolute_error: 0.5684\n",
      "Epoch 28/128\n",
      "72/72 [==============================] - 2s 23ms/step - loss: 0.4560 - root_mean_squared_error: 0.6753 - mean_absolute_error: 0.5315 - val_loss: 0.5094 - val_root_mean_squared_error: 0.7137 - val_mean_absolute_error: 0.5668\n",
      "Epoch 29/128\n",
      "72/72 [==============================] - 2s 23ms/step - loss: 0.4512 - root_mean_squared_error: 0.6717 - mean_absolute_error: 0.5285 - val_loss: 0.5070 - val_root_mean_squared_error: 0.7121 - val_mean_absolute_error: 0.5659\n",
      "Epoch 30/128\n",
      "72/72 [==============================] - 2s 23ms/step - loss: 0.4463 - root_mean_squared_error: 0.6680 - mean_absolute_error: 0.5258 - val_loss: 0.5028 - val_root_mean_squared_error: 0.7091 - val_mean_absolute_error: 0.5629\n",
      "Epoch 31/128\n",
      "72/72 [==============================] - 2s 23ms/step - loss: 0.4412 - root_mean_squared_error: 0.6642 - mean_absolute_error: 0.5227 - val_loss: 0.5001 - val_root_mean_squared_error: 0.7072 - val_mean_absolute_error: 0.5619\n",
      "Epoch 32/128\n",
      "72/72 [==============================] - 2s 23ms/step - loss: 0.4367 - root_mean_squared_error: 0.6608 - mean_absolute_error: 0.5203 - val_loss: 0.4990 - val_root_mean_squared_error: 0.7064 - val_mean_absolute_error: 0.5617\n",
      "Epoch 33/128\n",
      "72/72 [==============================] - 2s 23ms/step - loss: 0.4323 - root_mean_squared_error: 0.6575 - mean_absolute_error: 0.5177 - val_loss: 0.4995 - val_root_mean_squared_error: 0.7068 - val_mean_absolute_error: 0.5627\n",
      "Epoch 34/128\n",
      "72/72 [==============================] - 2s 23ms/step - loss: 0.4284 - root_mean_squared_error: 0.6545 - mean_absolute_error: 0.5154 - val_loss: 0.4960 - val_root_mean_squared_error: 0.7043 - val_mean_absolute_error: 0.5609\n",
      "Epoch 35/128\n",
      "72/72 [==============================] - 2s 23ms/step - loss: 0.4245 - root_mean_squared_error: 0.6516 - mean_absolute_error: 0.5131 - val_loss: 0.4932 - val_root_mean_squared_error: 0.7023 - val_mean_absolute_error: 0.5595\n",
      "Epoch 36/128\n",
      "72/72 [==============================] - 2s 23ms/step - loss: 0.4207 - root_mean_squared_error: 0.6486 - mean_absolute_error: 0.5108 - val_loss: 0.4905 - val_root_mean_squared_error: 0.7004 - val_mean_absolute_error: 0.5578\n",
      "Epoch 37/128\n",
      "72/72 [==============================] - 2s 23ms/step - loss: 0.4166 - root_mean_squared_error: 0.6455 - mean_absolute_error: 0.5084 - val_loss: 0.4859 - val_root_mean_squared_error: 0.6971 - val_mean_absolute_error: 0.5549\n",
      "Epoch 38/128\n",
      "72/72 [==============================] - 2s 23ms/step - loss: 0.4127 - root_mean_squared_error: 0.6424 - mean_absolute_error: 0.5060 - val_loss: 0.4823 - val_root_mean_squared_error: 0.6945 - val_mean_absolute_error: 0.5528\n",
      "Epoch 39/128\n",
      "72/72 [==============================] - 2s 24ms/step - loss: 0.4091 - root_mean_squared_error: 0.6396 - mean_absolute_error: 0.5037 - val_loss: 0.4782 - val_root_mean_squared_error: 0.6915 - val_mean_absolute_error: 0.5504\n",
      "Epoch 40/128\n",
      "72/72 [==============================] - 2s 23ms/step - loss: 0.4053 - root_mean_squared_error: 0.6366 - mean_absolute_error: 0.5013 - val_loss: 0.4771 - val_root_mean_squared_error: 0.6908 - val_mean_absolute_error: 0.5500\n",
      "Epoch 41/128\n",
      "72/72 [==============================] - 2s 23ms/step - loss: 0.4019 - root_mean_squared_error: 0.6339 - mean_absolute_error: 0.4993 - val_loss: 0.4732 - val_root_mean_squared_error: 0.6879 - val_mean_absolute_error: 0.5475\n",
      "Epoch 42/128\n",
      "72/72 [==============================] - 2s 23ms/step - loss: 0.3983 - root_mean_squared_error: 0.6311 - mean_absolute_error: 0.4968 - val_loss: 0.4667 - val_root_mean_squared_error: 0.6831 - val_mean_absolute_error: 0.5436\n",
      "Epoch 43/128\n",
      "72/72 [==============================] - 2s 24ms/step - loss: 0.3946 - root_mean_squared_error: 0.6281 - mean_absolute_error: 0.4945 - val_loss: 0.4645 - val_root_mean_squared_error: 0.6815 - val_mean_absolute_error: 0.5420\n",
      "Epoch 44/128\n",
      "72/72 [==============================] - 2s 25ms/step - loss: 0.3916 - root_mean_squared_error: 0.6258 - mean_absolute_error: 0.4928 - val_loss: 0.4632 - val_root_mean_squared_error: 0.6806 - val_mean_absolute_error: 0.5411\n",
      "Epoch 45/128\n",
      "72/72 [==============================] - 2s 24ms/step - loss: 0.3884 - root_mean_squared_error: 0.6232 - mean_absolute_error: 0.4908 - val_loss: 0.4640 - val_root_mean_squared_error: 0.6811 - val_mean_absolute_error: 0.5417\n",
      "Epoch 46/128\n",
      "72/72 [==============================] - 2s 24ms/step - loss: 0.3848 - root_mean_squared_error: 0.6203 - mean_absolute_error: 0.4884 - val_loss: 0.4595 - val_root_mean_squared_error: 0.6779 - val_mean_absolute_error: 0.5390\n",
      "Epoch 47/128\n",
      "72/72 [==============================] - 2s 24ms/step - loss: 0.3813 - root_mean_squared_error: 0.6175 - mean_absolute_error: 0.4865 - val_loss: 0.4606 - val_root_mean_squared_error: 0.6787 - val_mean_absolute_error: 0.5398\n",
      "Epoch 48/128\n",
      "72/72 [==============================] - 2s 24ms/step - loss: 0.3784 - root_mean_squared_error: 0.6152 - mean_absolute_error: 0.4846 - val_loss: 0.4582 - val_root_mean_squared_error: 0.6769 - val_mean_absolute_error: 0.5382\n",
      "Epoch 49/128\n",
      "72/72 [==============================] - 2s 24ms/step - loss: 0.3751 - root_mean_squared_error: 0.6125 - mean_absolute_error: 0.4827 - val_loss: 0.4577 - val_root_mean_squared_error: 0.6765 - val_mean_absolute_error: 0.5379\n",
      "Epoch 50/128\n",
      "72/72 [==============================] - 2s 24ms/step - loss: 0.3721 - root_mean_squared_error: 0.6100 - mean_absolute_error: 0.4807 - val_loss: 0.4582 - val_root_mean_squared_error: 0.6769 - val_mean_absolute_error: 0.5381\n",
      "Epoch 51/128\n",
      "72/72 [==============================] - 2s 24ms/step - loss: 0.3691 - root_mean_squared_error: 0.6076 - mean_absolute_error: 0.4789 - val_loss: 0.4570 - val_root_mean_squared_error: 0.6760 - val_mean_absolute_error: 0.5374\n",
      "Epoch 52/128\n",
      "72/72 [==============================] - 2s 24ms/step - loss: 0.3659 - root_mean_squared_error: 0.6049 - mean_absolute_error: 0.4768 - val_loss: 0.4568 - val_root_mean_squared_error: 0.6759 - val_mean_absolute_error: 0.5376\n",
      "Epoch 53/128\n",
      "72/72 [==============================] - 2s 24ms/step - loss: 0.3630 - root_mean_squared_error: 0.6025 - mean_absolute_error: 0.4748 - val_loss: 0.4573 - val_root_mean_squared_error: 0.6763 - val_mean_absolute_error: 0.5378\n",
      "Epoch 54/128\n",
      "72/72 [==============================] - 2s 24ms/step - loss: 0.3596 - root_mean_squared_error: 0.5996 - mean_absolute_error: 0.4726 - val_loss: 0.4591 - val_root_mean_squared_error: 0.6775 - val_mean_absolute_error: 0.5395\n",
      "Epoch 55/128\n",
      "72/72 [==============================] - 2s 24ms/step - loss: 0.3573 - root_mean_squared_error: 0.5978 - mean_absolute_error: 0.4711 - val_loss: 0.4588 - val_root_mean_squared_error: 0.6773 - val_mean_absolute_error: 0.5391\n",
      "Epoch 56/128\n",
      "72/72 [==============================] - 2s 24ms/step - loss: 0.3542 - root_mean_squared_error: 0.5952 - mean_absolute_error: 0.4691 - val_loss: 0.4559 - val_root_mean_squared_error: 0.6752 - val_mean_absolute_error: 0.5371\n",
      "Epoch 57/128\n",
      "72/72 [==============================] - 2s 24ms/step - loss: 0.3507 - root_mean_squared_error: 0.5922 - mean_absolute_error: 0.4667 - val_loss: 0.4573 - val_root_mean_squared_error: 0.6763 - val_mean_absolute_error: 0.5382\n",
      "Epoch 58/128\n",
      "72/72 [==============================] - 2s 24ms/step - loss: 0.3480 - root_mean_squared_error: 0.5899 - mean_absolute_error: 0.4649 - val_loss: 0.4584 - val_root_mean_squared_error: 0.6771 - val_mean_absolute_error: 0.5386\n",
      "Epoch 59/128\n",
      "72/72 [==============================] - 2s 24ms/step - loss: 0.3455 - root_mean_squared_error: 0.5878 - mean_absolute_error: 0.4631 - val_loss: 0.4587 - val_root_mean_squared_error: 0.6773 - val_mean_absolute_error: 0.5389\n",
      "Epoch 60/128\n",
      "72/72 [==============================] - 2s 24ms/step - loss: 0.3437 - root_mean_squared_error: 0.5862 - mean_absolute_error: 0.4618 - val_loss: 0.4605 - val_root_mean_squared_error: 0.6786 - val_mean_absolute_error: 0.5401\n",
      "Epoch 61/128\n",
      "72/72 [==============================] - 2s 24ms/step - loss: 0.3408 - root_mean_squared_error: 0.5838 - mean_absolute_error: 0.4599 - val_loss: 0.4591 - val_root_mean_squared_error: 0.6776 - val_mean_absolute_error: 0.5389\n",
      "Epoch 62/128\n",
      "72/72 [==============================] - 2s 24ms/step - loss: 0.3382 - root_mean_squared_error: 0.5815 - mean_absolute_error: 0.4581 - val_loss: 0.4549 - val_root_mean_squared_error: 0.6745 - val_mean_absolute_error: 0.5361\n",
      "Epoch 63/128\n",
      "72/72 [==============================] - 2s 24ms/step - loss: 0.3356 - root_mean_squared_error: 0.5793 - mean_absolute_error: 0.4564 - val_loss: 0.4530 - val_root_mean_squared_error: 0.6731 - val_mean_absolute_error: 0.5355\n",
      "Epoch 64/128\n",
      "72/72 [==============================] - 2s 24ms/step - loss: 0.3326 - root_mean_squared_error: 0.5767 - mean_absolute_error: 0.4544 - val_loss: 0.4510 - val_root_mean_squared_error: 0.6716 - val_mean_absolute_error: 0.5346\n",
      "Epoch 65/128\n",
      "72/72 [==============================] - 2s 24ms/step - loss: 0.3305 - root_mean_squared_error: 0.5749 - mean_absolute_error: 0.4529 - val_loss: 0.4476 - val_root_mean_squared_error: 0.6690 - val_mean_absolute_error: 0.5324\n",
      "Epoch 66/128\n",
      "72/72 [==============================] - 2s 24ms/step - loss: 0.3282 - root_mean_squared_error: 0.5729 - mean_absolute_error: 0.4512 - val_loss: 0.4447 - val_root_mean_squared_error: 0.6669 - val_mean_absolute_error: 0.5306\n",
      "Epoch 67/128\n",
      "72/72 [==============================] - 2s 24ms/step - loss: 0.3260 - root_mean_squared_error: 0.5710 - mean_absolute_error: 0.4500 - val_loss: 0.4392 - val_root_mean_squared_error: 0.6627 - val_mean_absolute_error: 0.5273\n",
      "Epoch 68/128\n",
      "72/72 [==============================] - 2s 24ms/step - loss: 0.3241 - root_mean_squared_error: 0.5693 - mean_absolute_error: 0.4485 - val_loss: 0.4410 - val_root_mean_squared_error: 0.6641 - val_mean_absolute_error: 0.5291\n",
      "Epoch 69/128\n",
      "72/72 [==============================] - 2s 24ms/step - loss: 0.3216 - root_mean_squared_error: 0.5671 - mean_absolute_error: 0.4466 - val_loss: 0.4350 - val_root_mean_squared_error: 0.6596 - val_mean_absolute_error: 0.5251\n",
      "Epoch 70/128\n",
      "72/72 [==============================] - 2s 24ms/step - loss: 0.3186 - root_mean_squared_error: 0.5645 - mean_absolute_error: 0.4443 - val_loss: 0.4312 - val_root_mean_squared_error: 0.6567 - val_mean_absolute_error: 0.5225\n",
      "Epoch 71/128\n",
      "72/72 [==============================] - 2s 24ms/step - loss: 0.3160 - root_mean_squared_error: 0.5621 - mean_absolute_error: 0.4425 - val_loss: 0.4340 - val_root_mean_squared_error: 0.6588 - val_mean_absolute_error: 0.5243\n",
      "Epoch 72/128\n",
      "72/72 [==============================] - 2s 24ms/step - loss: 0.3132 - root_mean_squared_error: 0.5597 - mean_absolute_error: 0.4405 - val_loss: 0.4338 - val_root_mean_squared_error: 0.6586 - val_mean_absolute_error: 0.5247\n",
      "Epoch 73/128\n",
      "72/72 [==============================] - 2s 24ms/step - loss: 0.3116 - root_mean_squared_error: 0.5582 - mean_absolute_error: 0.4391 - val_loss: 0.4332 - val_root_mean_squared_error: 0.6582 - val_mean_absolute_error: 0.5234\n",
      "Epoch 74/128\n",
      "72/72 [==============================] - 2s 24ms/step - loss: 0.3094 - root_mean_squared_error: 0.5562 - mean_absolute_error: 0.4376 - val_loss: 0.4374 - val_root_mean_squared_error: 0.6614 - val_mean_absolute_error: 0.5259\n",
      "Epoch 75/128\n",
      "72/72 [==============================] - 2s 24ms/step - loss: 0.3069 - root_mean_squared_error: 0.5540 - mean_absolute_error: 0.4358 - val_loss: 0.4399 - val_root_mean_squared_error: 0.6632 - val_mean_absolute_error: 0.5281\n",
      "Epoch 76/128\n",
      "72/72 [==============================] - 2s 24ms/step - loss: 0.3054 - root_mean_squared_error: 0.5526 - mean_absolute_error: 0.4348 - val_loss: 0.4411 - val_root_mean_squared_error: 0.6642 - val_mean_absolute_error: 0.5287\n",
      "Epoch 77/128\n",
      "72/72 [==============================] - 2s 24ms/step - loss: 0.3039 - root_mean_squared_error: 0.5513 - mean_absolute_error: 0.4336 - val_loss: 0.4404 - val_root_mean_squared_error: 0.6636 - val_mean_absolute_error: 0.5280\n",
      "Epoch 78/128\n",
      "70/72 [============================>.] - ETA: 0s - loss: 0.3018 - root_mean_squared_error: 0.5493 - mean_absolute_error: 0.4317Restoring model weights from the end of the best epoch: 70.\n",
      "72/72 [==============================] - 2s 24ms/step - loss: 0.3026 - root_mean_squared_error: 0.5501 - mean_absolute_error: 0.4325 - val_loss: 0.4425 - val_root_mean_squared_error: 0.6652 - val_mean_absolute_error: 0.5294\n",
      "Epoch 78: early stopping\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "WARNING:absl:Found untraced functions such as lstm_cell_layer_call_fn, lstm_cell_layer_call_and_return_conditional_losses, lstm_cell_1_layer_call_fn, lstm_cell_1_layer_call_and_return_conditional_losses while saving (showing 4 of 4). These functions will not be directly callable after loading.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "INFO:tensorflow:Assets written to: /Users/serhatsoyer/Repos/py4ML/datasets/time_series/cases/0_1_0_3_128_1_0_0_0_0/model/assets\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "INFO:tensorflow:Assets written to: /Users/serhatsoyer/Repos/py4ML/datasets/time_series/cases/0_1_0_3_128_1_0_0_0_0/model/assets\n"
     ]
    }
   ],
   "source": [
    "model, df, train_time = fn.get_model(Train_Sets, X_train, X_val, y_train, y_val)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "1c649667",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model: \"0_1_0_3_128_1_0_0_0_0\"\n",
      "_________________________________________________________________\n",
      " Layer (type)                Output Shape              Param #   \n",
      "=================================================================\n",
      " lstm1 (LSTM)                (128, 32, 32)             4608      \n",
      "                                                                 \n",
      " lstm2 (LSTM)                (128, 32)                 8320      \n",
      "                                                                 \n",
      " dense3 (Dense)              (128, 32)                 1056      \n",
      "                                                                 \n",
      " dense4 (Dense)              (128, 32)                 1056      \n",
      "                                                                 \n",
      " dense5 (Dense)              (128, 32)                 1056      \n",
      "                                                                 \n",
      " drop0.0625 (Dropout)        (128, 32)                 0         \n",
      "                                                                 \n",
      " dense6 (Dense)              (128, 32)                 1056      \n",
      "                                                                 \n",
      " out (Dense)                 (128, 1)                  33        \n",
      "                                                                 \n",
      "=================================================================\n",
      "Total params: 17,185\n",
      "Trainable params: 17,185\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n"
     ]
    }
   ],
   "source": [
    "model.summary()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "6a3689ef",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Total training time:          2.3 min\n",
      "Num of epochs:                78\n",
      "Best val RMSE epoch:          70\n",
      "Training time per epoch:      1.78 sec\n"
     ]
    }
   ],
   "source": [
    "fn.train_info(df, train_time)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "9939f067",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fn.plot_learn_curve(df)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "fc576e3a",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "2022-12-11 22:59:10.194996: I tensorflow/core/grappler/optimizers/custom_graph_optimizer_registry.cc:114] Plugin optimizer for device_type GPU is enabled.\n",
      "2022-12-11 22:59:10.259358: I tensorflow/core/grappler/optimizers/custom_graph_optimizer_registry.cc:114] Plugin optimizer for device_type GPU is enabled.\n",
      "2022-12-11 22:59:10.338210: I tensorflow/core/grappler/optimizers/custom_graph_optimizer_registry.cc:114] Plugin optimizer for device_type GPU is enabled.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "y_train_pred shape: (9216,)\n",
      "y_val_pred shape:   (3072,)\n",
      "y_test_pred shape:  (3072,)\n",
      "Common data type:   <class 'numpy.float32'>\n"
     ]
    }
   ],
   "source": [
    "y_train_pred, y_val_pred, y_test_pred = fn.get_predictions(Train_Sets, model, X_train, X_val, X_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "8c28b462",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "mse_train:     0.33\n",
      "mse_val:       0.43\n",
      "mse_test:      0.42\n",
      "rmse_train:    0.57\n",
      "rmse_val:      0.66\n",
      "rmse_test:     0.64\n",
      "mae_train:     0.46\n",
      "mae_val:       0.52\n",
      "mae_test:      0.51\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAj4AAAHFCAYAAADyj/PrAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjUuMywgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy/NK7nSAAAACXBIWXMAAA9hAAAPYQGoP6dpAACoYElEQVR4nOzdeXgT5fYH8O9M9jRJk7SULkDZZC07Cq5QEUEUF/QK3quCeJXFK/5AEcWlonIRVNTrhqLgdt1RULksAgUVAWVRWnbpQilt6ZK0afZk3t8faYamSaFL0kB7Ps8zj81kZvKmtc3hzHnfwwFgIIQQQghpA/hoD4AQQgghpKVQ4EMIIYSQNoMCH0IIIYS0GRT4EEIIIaTNoMCHEEIIIW0GBT6EEEIIaTMo8CGEEEJIm0GBDyGEEELaDAp8CCGEENJmUOBDCAk7xliDthEjRjT7tVQqFTIyMhp8rdTU1IAxuFwulJWV4bfffsPSpUvRp0+fFhsLIaTlUeBDCAm74cOHB2xr166FzWYL2r93795mv5ZarcYzzzyDkSNHNuq8//znPxg+fDhGjBiBu+66C6tXr8aNN96IP//8E4888kiLjoUQ0nKk0R4AIaT12bVrV8Dj0tJSCIIQtD+aTpw4ETCedevWYenSpfjmm2/w4osvIjs7G+vXr4/iCAkhkUAZH0JIVMhkMjzxxBM4dOgQHA4HTp8+jRUrViA+Pj7guPT0dGRmZqKsrAw2mw35+fn4+uuvoVKpkJqairKyMgDAM888I96+WrlyZZPG5HA4cO+998LlcmHu3Lni/vj4eLz55ps4cOAALBYLSkpKsHnzZlxxxRXiMecaS7du3bBixQocPXoUVqsVJ0+exHfffYe0tLQmjZUQ0jSU8SGEtDiO47BmzRpceeWVWLJkCX799VekpqZiwYIF2Lp1K4YOHQqHw4HU1FSsXbsWP//8M6ZOnQqz2YyUlBSMHTsWcrkcRUVFGDNmDDZs2ID33nsP7733HgBfhqmpioqKsGfPHlx22WWQSCTwer0wGo0AgAULFqC4uBgajQa33HILtm7dilGjRmHbtm3nHEtycjLKy8vx2GOPobS0FEajEZMnT8auXbswaNAgHD16tJnfVUJIQzHaaKONtkhuK1euZBaLRXw8ceJExhhjt9xyS8BxQ4YMYYwxNn36dAaATZgwgTHGWP/+/eu9dlxcHGOMsYyMjAaNJTU1lTHG2MMPP1zvMZ999hljjLF27dqFfJ7neSaRSNiPP/7IVq1a1aSx8DzPpFIpO3LkCHv55Zej/jOijba2stGtLkJIi7vhhhtgMpnw/fffQyKRiNsff/yBoqIisTj4jz/+gNPpxLvvvou7774bXbp0aZHxcRwXtG/atGnYs2cP7HY7vF4vPB4PrrnmGvTu3btB15RIJHj88cdx4MABOJ1OeL1euN1u9OjRo8HXIIQ0HwU+hJAW1759exgMBrjdbng8noAtKSlJrPPJycnBNddcg9OnT+PNN99ETk4O/vrrL8yaNSui40tNTYXD4UBFRQUAYPbs2Vi2bBl27dqFW2+9FcOGDcPQoUOxbt06qFSqBl1z6dKleO6557B69WqMHz8el1xyCYYOHYo//vijwdcghDQf1fgQQlpcWVkZysrKMHbs2JDPWywW8etffvkFv/zyC3iex9ChQ/Hggw/itddeQ0lJCb744ouwjy05ORlDhgzBtm3b4PV6AQB33nknMjMzMXPmzIBjtVptg69755134qOPPsITTzwRsD8+Ph5ms7nZ4yaENAxlfAghLe6HH35AfHw8JBIJ9uzZE7SFKvQVBAG//fYbHnjgAQDA4MGDAQBOpxMAwpI1USqVeO+99yCTybBkyRJxP2NMfB2/fv364dJLLw3Yd7axhLrGuHHj0KFDh2aPmxDScJTxIYS0uM8//xz/+Mc/8L///Q+vvfYafvvtN7jdbnTo0AHp6elYs2YNVq9ejWnTpuHqq6/G2rVrceLECSiVSkydOhUAsGnTJgBAdXU18vLycNNNN2Hz5s2oqKhAWVkZ8vPzzzqGTp06YdiwYeB5HrGxsRg0aBCmTp2K1NRUzJkzBz/++KN47A8//ICnnnoKzzzzDLZt24aePXvi6aefRm5uLqTSM39GzzaWH374AVOmTMHhw4exf/9+DBkyBHPnzkVBQUEEvsOEkLOJeoU1bbTR1rq3urO6ADCJRMLmzJnD9u3bx2w2G6uqqmIHDx5kb7/9NuvWrRsDwIYNG8ZWrVrFcnNzmd1uZ6WlpSwzM5PdcMMNAde6+uqr2Z49e5jdbmeMMbZy5cp6x+Kf1eXndrtZeXk5+/3339nSpUtZ7969g86RyWRsyZIlrKCggNlsNrZ792524403spUrV7Lc3NwGjSU2NpYtX76cFRcXs+rqavbTTz+xyy+/nGVmZrLMzMyo/4xoo62tbFzNF4QQQgghrR7V+BBCCCGkzaDAhxBCCCFtBgU+hBBCCGkzKPAhhBBCSJtBgQ8hhBBC2gwKfAghhBDSZtAChiEkJycHLJlPCCGEkPOfVqvFqVOnznoMBT51JCcno7CwMNrDIIQQQkgTpKSknDX4ocCnDn+mJyUlhbI+hBBCyAVCq9WisLDwnJ/dFPjUw2KxUOBDCCGEtDJU3EwIIYSQNoMCH0IIIYS0GRT4EEIIIaTNoMCHEEIIIW0GBT6EEEIIaTEqnRYcx0Xt9SnwIYQQQkjEdR0yEHf8+2lkbP4eFw0fGrVx0HR2QgghhEREjEGPi28ch2G33oiELqni/mG33oSjO36Pypgo8CGEEEJI2HAch4uGD8WwW29C2tVXQSqTBR3TeWA/SGQyeN3uFh8fBT6EEEIIaTZtfBwuufkGDLt1POI6pIQ8pig7CyV7duLQtl8geDwtPEIfDgCLyiufp7RaLaqqqqDT6WjlZkIIIeQsOJ5Hr8uHY/htN6L3VZdDIg3Op1SWlOL0H79j3EAt+nWPg1wuhcvlwZEjhVj8wtfIzNwflrE09PObAp86KPAhhBBCzk6f2B7DJozHJbfcAH1i+6DnBa8Xh37egV2r1qC9pBpvvzUdWp0a5eVVcDrdUChkMBq1qLbYMX36W2EJfhr6+U23ugghhBByTrxUgr4jrsCw225Cz8uGgeeDJ4ZXFBZh17ff4/fVP6CypBQcx+Hl9Qug1alx8mSZeJzd7kJhYTlSUuIw77HbsHVrFhhrmTwMBT6EEEIIqVdcxw4Yfut4DL3peuji44Ke97o9OLD1Z+z8eg2O7vwdTBDE5wYN6oqePVNQXl4V8toVFRb07JmCQYO6Yu/e4xF7D7VR4EMIIYSQABKZDP1GjcDwW2+qd82d0vwC7PrmO+xe8z9YyitCHhMfr4NcLoXTGXr2ltPphtGoRXy8LmxjPxdawLCFrVy5EowxvP3220HPvfnmm2CMYeXKleK+du3aYdmyZcjPz4fD4UBRURHWr1+P4cOHi8fk5uaCMRa0zZs3L+zjv+qqq7B7927Y7XYcP34c06ZNa/C5RqMRBQUFYIwhNjZW3D9ixAisXr0ap06dQnV1Nfbt24e///3vYR87IYS0ZhzHYfDgbrj22kEYPLhbk1ZHTuiSihvnzkLG5u9w14vPBQU9HpcLe/+3EW9NfQCLx09E5opP6g16AKCsrAoulwcKRfCUdgBQKGRwuTwoKwudEYoEyvhEwYkTJzBp0iTMnj0bDocDAKBQKHDHHXcgPz8/4NhVq1ZBJpNh8uTJyMnJQfv27TFq1CgYjcaA45566iksX748YF+4i7M7d+6M//3vf1i+fDnuvPNOXH755XjrrbdQWlqKb7755pznv//++9i/fz86dOgQsP+yyy7D/v37sXjxYpSUlOD666/HRx99hKqqKvzwww9hfQ+EENIapaf3x7zHbkPPnimNnjUlUyowcMwoXH337Ujo0TPkMcXHc7Fr1XfY8/06WM2V4DgOgwZ1RXy8DmVlVdi3Lydkjc6+fTk4cqQQ/fqlorCwPOh5o1GLrKx87NuX07Q33gQU+ETB3r170bVrV0yYMAGffvopAGDChAkoKChATs6ZH35sbCyuvPJKjBgxAj/99BMAX9D0++/Bq11aLBaUlJREdNzTp0/HiRMnMHv2bADA4cOHMXToUDzyyCPnDHymT58OvV6PZ599FuPGjQt4btGiRQGPX3/9dYwZMwa33HILBT6EEBJC7cCjc5f2ePjhm6HVBs6a6tcvFcuWzQyaNeU/t0v/PogfcDG6XHY5ZGp10Gt4XS7sXfcjdn61Bnl/Zon7GxNkMcaw+IWvsWzZTKSkxKGiwhIwq8tisWPxC1+3WGEzQLe6omblypW45557xMdTp07FihUrAo6prq6GxWLBzTffDLlc3qzXy83NRUZGRrOucemll2Ljxo0B+zZs2IChQ4dCGmLtBr/evXvj6aefxt133w2hVtHb2cTGxqKiov70KSGEtFXp6f2xbv0CrPpmPj7+5GG89tr9SE1tj8pKG+x2FwSBibOmNFoV5j12m3jb65qxF+O/P72HBz95D5c9PB89rhkdFPRoYcOQmBKMizmCks3fBQU9y96ZicGDu4ExBrPZiupquxhkpaf3DxpvZuZ+TJ/+FrKy8qHRqJCUZIRGo0JWVj5mhGkqe2NQxidKPv74YyxatAipqalgjOHyyy/HpEmTMHLkSPEYr9eLKVOmYPny5Zg+fTr27t2Lbdu24fPPP0dWVlbA9RYvXoznn38+YN8NN9yAbdu2AQCOHz+OsrIyNEdiYmJQVqmkpAQymQzx8fEoLi4OOkcul+Ozzz7D3LlzUVBQgK5du57zdW699VZcfPHFjaofIoSQC1VDbxsBNYHHspmIqykalsulkMl8H+Xduyfh1KlylJSYxeMrKizo0SMF19x6DfqMHo0ul12OIvCAN/C6Mk5AImdCnL0QGmYF5wAMdaaacxyHxUumoFOnBACATqcGYwwOhxvFxSbodKp6p6ZnZu7H1q1ZDX6fkUSBT5SUl5dj7dq1mDx5MjiOw9q1a1FeHnz/85tvvsHatWtx5ZVX4tJLL8XYsWPx6KOP4p///Cc+/PBD8bgXX3wRH3zwQcC5hYWF4tfXXHNNWMZd939S/78i6vufd9GiRTh06BD++9//Nuj6I0aMwAcffID77rsPBw8ebN5gCSHkPFffbaMli1fBbK4OCBIAYN5jtyEuXge1WgGJhIfXK4Ax399gnueQnBwHm80Fi8UGNyQochtRZk/G2IxnAQTFO0hUudFX70DPWAc8TicOHbKKz9Wdan7ffWMwYEAXAIDH44Ug+F5TpZIjNbUdiotNZ52azhhrsSnrZ0OBTxStWLECb7zxBgDggQceqPc4p9OJTZs2YdOmTXjuueewfPlyLFiwICDwKSsrw/Hjkf0fqri4GImJiQH7EhIS4Ha7QwZtAHD11VejX79+uO222wCcCZTKysqwcOFCPPPMM+KxV111Fb7//nvMmTMHH3/8cWTeBCGEnCf82Zu6KxpfckkPrP1fBhwON1wuN2w2J44cKcQ33/wqBkgSCQ+XywOeP9OAwZ+V4eMSccTBo4yPg8AFZ3fkvICLNFb0jbUjScuJ58lUcqjVSthsvkk3taeacxyH++4fA57n4XC4xGsJAoMgeCCTSWEwaGGzOVp0anpTtKrAJyMjI+CDFPB9WCclJUVnQOewfv16sXZnw4YNDT7v4MGDuPnmmyM0qvrt2LED48ePD9h37bXXYvfu3fDU02zu1ltvhUqlEh9ffPHFWLlyJa688sqAQG3EiBH44YcfMG/evKDZaYQQ0tpwHId5j90WsKKxVqtCp04JUCp9U7/lcikYU0Kj8W19+3ZCTIwCMpkUbrcvmvEFHgwOQYJDZgUOmFUwuaSAJPg1k1Uu9NRUo4vaBimPmqDJ91r+jJEmRiEGPrWnmg8a1BXJyUZ4vQI4jgvK8nu9XqhUMthsjhadmt4UrSrwAYDs7OyA2zpeb93E3vlDEAT07t1b/Louo9GIr776CitWrMD+/fthsVgwdOhQPProo1izZk3AsVqtFu3bB/ZLsdls4pT2TZs24dtvv8Wbb77Z5PEuW7YM//rXv/Dyyy9j+fLluPTSS3HvvffijjvuEI+5+eabsWjRIvF91Z6lBgDx8fEAgEOHDqGyshKAL+hZu3YtXnvtNaxatUp8Hy6XCyaTqcnjJYSQ81XdFY21WhW6dEmETBYYsXAcB6VSDplMCoXCBblcCo7jIAgAx/E4aZPj0OkY5FqVEBC8bo8cHhzavAWy03/hwQW3wu32wP9x4w+afK0nagKZWpeoPdV89OiBAACHwwmlUgG3O/Afu4LAIJdLcOpURYtOTW+KVhf4eDyeiE/rDqezrbVTXV2NXbt2Yfbs2ejWrRtkMhkKCgqwfPly/Pvf/w449rnnnsNzzz0XsG/ZsmWYMWMGAKBbt25i0NFUeXl5GDduHF555RU88MADOHXqFGbNmhUwlT02Nha9evVq1HWnTJmCmJgYzJ8/H/Pnzxf3b926Fenp6c0aMyGEhIt/gcBhw3uCA4PZbENZWRVKSysbVKhbu4i5+0XJASsaJyYagoKe2iQSHmq1EgBg8/A4YNHggEmFSnfoc9rLrJCV/IV3nnsHWzbtw5Ah3eF+8mZIpRK4XGeCFrfbC4WCE7M4VqsLKpU8aKq5fyFCp9ONxEQZZDIpPB7vmdtkMgkEQcDy5RuiUrDcGK2qO3tGRgbmzp2LyspKOJ1O7Nq1C/Pnz0dubm6958jlcigUCvGxVqtFYWEhdWcnhBAiSk/vj8VLpqBv31TI5RJwnG81GEEQUFlpw549f511scDaRcwxMUrwHIcYjRKnTlXAbneie/fkswY+AgPyq+XINimRY5GDhcjuqCVedFdXQ1GWg9nTlmLv3uNiEMJxHHb99jIGDuwKQWDwes8UJ0ulvpoht9uDigpLyHV5OI7DuvUL0K9fKqqqbEhMNECplNe67cWQlZWPYZc8HLXAp6Hd2VtV4DN27Fio1WocPXoU7du3x5NPPolevXqhb9++9a4JE6ouCAAFPoQQ0so1dBp5enp/fPjRbCQmGgD4amPqtoNwu704fdqMyXe/EhT81J6CLpP5pqDzPAeJxBfolJdXwmDQQiIJXlqvysXjgFmJg2YlLCGyOxwYUjUupBns6Kh0oCC/JGjBwrrvo127WPiG7wsBGANKSyvx/HNfIC+vpN7vhf99aLQqVFRYwPM8VCo5YmKUMJmqo7ImT21tMvCpS61W4/jx41iyZAleeeWVkMdQxocQQtqehq4+7M90jBiRBp7nwPN8TVFwIMZ82Z8//sgJyHr4zx86tLs4Bd0/FVwq5SGXywKmowOAlwG5Fl92J69aDoTI7mikHqQZHOils0Mj9YIxX03r7Nnv4d131p/zfffrlwq5XAaXy42srPwGtbZozPctGhoa+LS6Gp/abDYbsrKycNFFF9V7jMvlgsvlqvd5QgghrUt908hDtXgYNKgr+vVLBc8HBid1cZyvDqdfv84YPLgb9uz5CwAwZEh3DBrUtSbokcDlOtOl3OMRwJgbcrkMHAeYXTyyTSocNCtg8wRnd3gwdNE60VNjRbLCDoVciqKiCpQ63fB6Bej1McjLPXuNa3MXEjyfFiJsqlYd+MjlcvTu3Rs///xztIdCCCHkPBBqGjkAscVDSp3VikeNGoC4OB2k0vrrb2qTy6W466507NnzF2bPuRkZGZOg1Z5pCSGRKODxeMTp6C6PgONWGbIqlChyKkNeM1buRZrejj56B2JkDIwxuFx8TTGyAzabEyqVvMFdzpu7kOD5shBhU7WqwOfFF1/E999/jxMnTiAhIQFPPvkkdDpdwEJ/hBBCWr/66nfqTiOvq/ZqxbGxMZjz8C1nLToOZfS1gzBnzi3496K7g+p2OA6QyaSocEqRbVLiWLUaDiH4+hKOoZvWiX4GBzrEuOGvxWGMA8BBLpfCYrHDZnMCiE6X8wtVq2pS2qFDB3z22Wc4cuQIvvnmG7hcLgwfPhwnTpyI9tBEK1euBGO+iN3tdiM/Px9vvfUW9Hp9wHG5ublgjGHixIlB18jOzgZjDJMnTxb3DRw4EN9//z1KSkpgt9uRm5uLzz//HHFxcQAg9gQLtQ0bNqxR7+Gqq67C7t27Ybfbcfz48Qb31Jo8eTL+/PNP2O12FBUV4fXXXw95XLdu3VBVVUVr+BBCmqR2E8+VH/wfVn0zH+vWL0B6en/E1/S48k8jr8vpdEMul6Jdu1gsXjIFcXHaRt/G6d49Cc8suAMSiW+VY0Hwne8WgINmBb7M1eOz/HbIqtIGBT2W4iJcJJzA1O6lGJtSiY4a3zjdbm/N2jlMLEw2m61QqeRISYmLSpfzC1WryvjUXkjvfLZu3Trcc889kEql6NOnD1asWAG9Xo+///3vAcedOHEC99xzD7744gtx37Bhw5CYmIjq6mpxX7t27bBp0yZ8//33GDNmDMxmM7p06YIbb7wRarU6oJ3EqFGjcODAgYDXqa/dRCidO3fG//73Pyxfvhx33nknLr/8crz11lsoLS0NWM+nrtmzZ+Phhx/G3LlzsWvXLiiVypANS6VSKT777DP8/PPPuOyyyxo8LkII4TgO9903Bk89NREKpRwlJaag+p2Xl66Gy+WBQiGD3R5c3+lfrTg+Xoe+fX3/YPR4BLERaENIpRJIpRK43R7wPI9ShxQHK1U4ZFbAJQTnGyQQoLUV4/npC5C7bz8effRWXPvcPyAIHBgTwPP+dXIYvF4BAOdbZVnjWxW/McXJpJUFPhcKp9MpLrJYWFiIL774AlOmTAk67r///S9mz56NDh064OTJkwCAqVOn4r///S/uvvtu8bjLLrsMOp0O//znP8WVqvPy8pCZmRl0zfLy8mYt8Dh9+nScOHECs2fPBgAcPnwYQ4cOxSOPPFJv4KPX6/H8889j/Pjx2LJli7g/VBPS559/HocPH8bmzZsp8CGENJh/ttGVV/aFXO5bXE8mk8BkqobL5YHJVA2DQYMpU0ahrKwKqakJOH3aDKlUAo/HG3TLSK+PgUwmgdcrNPpWFwC4vBwOV8XggFmFEocs5DFxCje6KSqhtxdh1mNvIHfffqSn98fMmePA877ZX16vULPWjgSMMRQVmVBd7YDRqMHC57/Azp1HLrji4mhrVbe6LkRdunTB2LFj4XYHp11LSkqwYcMG8ZaWSqXCxIkTsWLFioDjiouLIZPJcMsttzRrLP7bYSNGjKj3mEsvvRQbN24M2LdhwwYMHToUUmnoOHr06NHgeR4pKSk4ePAgCgoK8MUXX6BDhw4Bx6Wnp+Nvf/vbWRu2EkJIXf5ZWoMHd4NEwokzp7RaNTp1SkDXrom46KJktGsXi4svvgg9eqTAaNSgV6+O6No1Ed27J6NPn47o0qU9XC4PNm/6Ax06+MoEpFIeACe2dzgbxoBimxSbCjVYftSILcW6oKBHxgtIM9gxqasJEzuVwXp4H2bNeAOZmfvFwmu5Qgar1VHTSsLXDsK32jIHvV4Do1GDrKx8vP32uoBFCknDUMYnCm644QZYLBZIJBKxgac/g1LXihUr8PLLL2PhwoW47bbbcPz4cfz5558Bx+zatQsLFy7Ep59+imXLluG3337Dli1b8NFHH+H06dMBx/76669BfcFiY2MhCALcbjcOHz4Mm81W79gTExODMkYlJSWQyWSIj49HcXFx0Dldu3YFz/OYP38+HnroIVRWVuL555/Hjz/+iP79+8PtdsNoNOKDDz7AnXfeSesnEUIarPYsrYoKC2JjYyCX82LQAPimmftbKwCAUikTp6ZLJDwkEkAmk0CpVECplGPeY7dBEGo6nfNczde+a9W+jp/Dy+GwWYFskwplztAfq+2VbqQZ7LhI54RS6lvz5/RpC56Y/5E49b124bVUKkFqakJAawhBEBATo0BlZTXV8zQDBT5RkJmZiRkzZkCtVuOf//wnevToUW+h79q1a/HOO+/gqquuwtSpU4OyPX5PPvkkli5diquvvhrDhw/H9OnTMX/+fFx11VXIzs4Wj5s4cSIOHToUcK4/EDp16pTYXPRs6v6y+f8I1PdLyPM85HI5Zs2ahR9//BGArx6ruLgY6enp2LhxI5YvX45PP/2Ulh4ghDSKP1iwWh3Q6dSQSHhwIZba8bdW4Dj/ism+WVI+vv0cB6jVippFBc/8A7G+BQtP2aTINqlwrEoBDws+RsEL6KV3Is1gRzulN+DcoqIKSCQSxMVpxf21C6/tdhfy808HtYbwer148421VM/TDBT4RIHVasXx4741EB566CFs2bIFGRkZePrpp4OO9Xq9+Pjjj7FgwQIMGzbsrLezKioq8PXXX+Prr7/G448/jn379uGRRx4JqB8qKCgQX7spiouLkZiYGLAvISEBbre73iLpoqIiAIE1PWVlZSgrK0OnTp0AAFdffTVuvPFGPPLIIwAg/nFyu924//77sXLlyiaPmRDSeo0aNQCJiXoAHGQyacigx8//jzSO808L92dvAk/y3S5jkEhY0HN2D4dDZiWyzUpU1JPdSVa70FdvR89YF6QhCkoEwVcsLQgM5eVnMtz+RqD+wmuLxQ6LxQ61WgGpVAKJhIdUKsHmzX8GX5Q0GNX4nAcWLFiARx55BElJSSGfX7FiBUaOHIk1a9bAbDY36JputxvHjx9HTExMGEcK7NixA6NHjw7Yd+2112L37t3weDwhz9m+fTsAoGfPnuI+g8GA+Ph45OfnA/DVDg0cOFDcnn76aVRVVWHgwIH49ttvw/oeCCEXPo7jcP+0sXj4kVsglUpr+mc15nwE3bKqfW3fasq+j0jGgHyLDP8r0GL50Tj8VKIJCnpUEgGD42y4s2s5bu9SiT56JyQcxHYUAOD1CvB6BUgkPBIS9NDrY7Dw33cjPb0/AGDfvhwcOVIIo1EbcG2bzYmqKhvUagWOHCmktXqaiTI+54Ft27bhwIEDmD9/Ph588MGg5w8fPoy4uLh6a2+uv/56TJo0CZ9//jmOHj0KjuMwfvx4jBs3Dvfcc0/AsXFxcWjfvn3APrPZDKfTieTkZGzevBl33303fv/995CvtWzZMvzrX//Cyy+/jOXLl+PSSy/FvffeG7CUwM0334xFixaJt82OHTuG1atX47XXXsP999+PqqoqLFq0CIcPHxZnnh0+fDjgdYYOHQpBEIKm3hNCLkwNaQh6tmNqP9e5c3tMuPUyXHllH7HdQ0NXVm74eAGrm8cBswIHzCpUukJfP1FuQ5rBgZ4Gb012h0EQUDMN3gO5XFpzm8p/y8xfGgCYTNVBbTIWv/A1li2biZSUOFRUWMTp+EajltbqCRMKfM4TS5cuxcqVK7F48WJx6npt9XWXB3y3kGw2G15++WV07NgRTqcTx44dwz//+U988sknAcdu3rw56PxJkybhiy++gEwmQ69evaBWq4OO8cvLy8O4cePwyiuv4IEHHsCpU6cwa9asgKnssbGx6NWrV8B5d999N1555RWsXbsWgiBg27ZtGDt2bL1ZIkJI69GQxpb1HfPNN78iVqfGLbdciuRkI2L1MYiJUYr1Or4FByVip/PmEhiQX+1rEJpjkYOFaBCqlnrRJ9aB3jo7KgoLUZVvg1PZAZxSDsA3A0smk8LrFeB0eqBQSMVbar46HV99UXW1A1VVtoA2GZmZ+zF9+lvi98Jo1MLl8tBaPWHUqruzN0VDu7sSQgg5t/oaghqNWlRb7Jg+/S0ACDomMdGAdu1iaxUW+2ZXSSRcwC0qj8cLiUTSqNtcoVjcPA6YlDhgVsLiDg6iODCkalxI0zvQRecCD4AxAXl5p1FZaYXRqEWHDvEAgNOnzYiL09XMHOPFGWReryBma3iex/HjRWKfLY1GhVsn/FvsgdWQDBkJRN3ZCSGERFXdqeYajQoajQperxcWix1GoxbzHrsNHACDUSse07lze6hUioBghjFAKg2cQcpxXLNucXkZkGvxZXfyq0Nnd7QyL/rqHeijt0MjFQLG5PUySKUSJCUZ4XJ58Mcfvtqbdu1iAfies1p99Tnt2+trVl329eqy213ioolOpxtGoxbx8bpa7/fCbgR6PqPAhxBCSEQMGtQVAwZ0hk6nrpO98QUyXq8XI0emwX/zQa+PCWrq6Vc74KivKLmhKl08sk0qHDArYPOEzu501bqQZnAgVeMCz6GmtyHEWh2PxwuTqRpPzP8IZWVVYlbG/75HjRqAB/51PWQyKVwuNxhjkEh86wsJgoDi4jO9CP1tMhrSWZ00HwU+hBBCImLUqAGIj9cFLCboV7cg2ePxhlwvJ1w8AnDcokC2SYkCqzzkMbEyL9IMdvTROxAjO3NbyX+HqfaaZUVFFZDJpCgrq8LGjfsCrrN373Hs3Xscu3f/Jdbq1M4AFRebYLGcmaxCndVbFgU+hBBCwo7jONxyy6XilPBzCfesLL8KpwTZJiUOmZWwe0M0COUYummdSDM40DHGHVQrxJivWPnM4oc8bDbfGjsajeqsWZrMzP3YujUrKAPk8XjA8xzN1ooSCnwIIYScU2OLbQcN6orUzgnwzZ+JXCYnFI8AHK3ytZA4ZQvdINQg96CfwYHeegdU0vrfh78jOs9z4HkeXq/vNlVDszT+Wp26GSCarRU9tIBhC1u5ciUYY3j77beDnnvzzTfBGAu5SvGll14Kj8eDdevWBT3nby4aahs2bFjY38OMGTOQk5MDu92O3bt344orrjjr8SNGjAg5ttoLGvbp0wdff/01cnNzwRjDQw89FPZxE0KaJj29P9atX4BV38zHyg/+D6u+mY9fti/B3LkTMHhwt5A1N/HxOsTEKJtdj9MYpQ4JMos0WH4kDhsLdUFBj4Rj6B3rwG2dTbirmwmD4+31Bj0ejxdHjxaCMQalUg6JhIfd7kJxsQk6napJWZrMzP24bmwGbp3wb9wz5VXcOuHfuG5sBgU9LYwyPlFw4sQJTJo0CbNnz4bD4QAAKBQK3HHHHeJKxnVNnToVr7/+Ov75z3+iY8eOKCgoCDpm1KhRQQv+1ddGoqluv/12vPrqq5g5cya2b9+OadOmYd26dejTp0/IMdXWo0cPVFWdSQuXlpaKX6vVauTk5OCrr77CK6+8EtYxE0Karu50dJlMiuRkI1JS4nDxxRdhTrkFf/6ZG5S1KC+3QC73BR6hGnuGi8sLHK1SIsukRIk9dHYnXuFBmsGO7hoblBKGigoLvHIteF4S1IAU4OBwuPD00//Fq6+swX33j8F9941BcrIRgC/z1ZwsDc3Wij4KfKJg79696Nq1KyZMmIBPP/0UADBhwgQUFBQgJyc4bapWq3H77bfj4osvRmJiIqZMmYLnnnsu6Ljy8vKgzunhNmfOHLz//vt4//33Afi6yo8ZMwYzZszA/Pnzz3ru6dOnUVlZGfK53bt3Y/fu3QCAF154IbyDJoScVd3bWH/8kYuBA7ugXbtYPJ1xhzgdXadTIyFBD57n4HJ5IJVKoFYrMHhwN3z08Rw8//zn2P37X4iL06JdOx0YEyIS9DAGlDikyDYpcaRSAbcQfPNCxjP01DmQZnCgvcoDxgR4vb6xCALD1q1ZOHAgH1ddlYbkZKOvd5ZXQE5OMZ588hNs2eLrh/XuO+ux/N0NtKZOK0KBT5SsXLkS99xzjxj4+Duvjxw5MujYiRMn4siRIzh69Cg++eQTvP766yEDn7NJTU1FXl4eRo4ciW3btjVpzDKZDEOGDAkKTDZu3IjLLrvsnOfv27cPSqUSBw8exPPPP4+tW7c2aRyEkPCpu2Iyz/saBPt7ShkMMWDMN/NIKuXBcZxY88IYQ0yMAiqVr3v4G29Mh8vlhc3mBMf51qsJZ+Dj8HI4YlYgy6xCmSP0x1d7lRtpegd6xjohl/iCE4/H1xnd4XDi2LEiTJ/2JvbuPS6O7VxBDWVpWhcKfKLk448/xqJFi8T6nMsvvxyTJk0KGfjce++9YuuJ9evXQ6PRYNSoUUHtJ3799VcIghCwLzY2FoIgwO124/Dhw/X2+2qI+Ph4SKXSoKxSSUlJUMf22oqKinDfffdhz549UCgUuOuuu7B582aMHDkSP//8c5PHQwhpmPo+3EPdxkpNTYBUKoHH40VVlU2cis7zvmDAt3oyH7DejkRyZnVluZyDQiEF0LimofVhDDhlkyLbpMKxKgU8LPiicl5Ab70TaQY72im9tc71rb3jX9lZrfa1utDp1GJwQ0FN20OBT5SUl5dj7dq1mDx5MjiOw9q1a0PW4/To0QOXXHIJJkyYAMC34NcXX3yBqVOnBgU+EydOxKFDhwL2+QOhU6dOiU1DmytUY8GzpX2PHj2Ko0ePio937tyJjh074pFHHqHAh5AIq68H1pLFq/DovFuh1alx8mQZAKBDh3jwvK/GRSaTwmDQAAis0TlXMBOutXjsHg6HzEpkm5VBndD9ktVupBnsuEjnhKzW3S632wOvV4BCIQfPnwmAzGYrOnduj2XvzMT0aW9RUXEbRYFPFK1YsQJvvPEGAOCBBx4Iecy9994LmUyGwsJCcR/HcXC73dDr9TCbzeL+goICHD8euX+5lJWVwePxBGV3EhISGl1btHPnTtx5553hHB4hpI76+mT165eKFStnQaGQievQqNUKKJVyuN2+jIkgCJDJZDXFv1xEC5T9GANOWmXIMilx3KKAN0R2RykR0Efvq90xKrwhrgK43V54PAKkUgGMCZBKpQGZHq1WhcVLpmDYJQ9TrU4bRNPZo2j9+vWQy+WQy+XYsGFD0PMSiQR333035syZg4EDB4rbgAEDkJ+fj3/84x8tOl632409e/Zg9OjRAftHjx6NX3/9tVHXGjRoEIqKisI5PEJILbX7ZJ08WQa73QVBYLDbXSgsLIdGo4JeHwOn0w3At4Cgb60aLiBr4/V6AUQ26LG6OfxeqsIHfxmxKl+Po1XKoKCnY4wL13Wowj97lOOqRGu9QQ9jgFwuhUolB2MCZDIpOA7weDxwuz0QBF990oABXXDffWMi9p7I+YsyPlEkCIJ4+6lubQ4A3HDDDTAYDHj//fcDpoEDwNdff417770Xb775prgvLi4O7du3DzjObDbD6XQiOTkZmzdvxt13343ff/+9yWNeunQpPv74Y+zevRs7duzA/fffj06dOmHZsmXiMf/+97+RkpKCyZMnAwAeeugh5OXl4cCBA5DL5bjzzjtx2223ibfvAF/hdJ8+fQAAcrkcKSkpGDBgAKqrqyOaxSKktahbx8NxHHr2TEF5eeiVhW02J2JjfT20bDYnkpKMkEgkkEgkABgCEyHhD3oEBuRX+xqE5lrkEEK8hloqoE+sA30NdhgUwX8j6yOR8GDMd1uOMQan0yP+jRUE30rMSqUc990/BsuXb6CsTxtDgU+UWSyWep+79957sWnTpqCgBwBWrVqFJ554AoMGDUJFRQUABNX8AMCkSZPwxRdfQCaToVevXlCr1c0a75dffom4uDg8/fTTSEpKQnZ2NsaNG4cTJ06IxyQlJaFTp07iY7lcjpdeegkpKSmw2+04cOAAxo0bF7AYY3JyMv744w/x8dy5czF37lxs3boV6enpzRozIa1dqDqesrIqaDTKoMBHq1UjMdEApVIGnueRnBwnBghnkjpnCpOlUmlYipT9LG4eB0xKHDArYXGHalPB0FnjaxDaResCxxp/q81isYMxICZGDo9HgCAE1yV6vQKSk40YNKgrFTe3Mb6WuESk1WpRVVUFnU531qCEEELOB/XV8bRvb0BsrBqFheViHU+7drFISjKA533T0iPZFLQ2LwPyLL7sTl61HCxEdkcj9aKvwYG+egd08jPZnabUFmVn58Pl8mDgwK5wOFxBz8tkUjgcTlitTtwz5dWgJqPkwtTQz2/K+BBCyAWqbh2Pn93uQl5eCfr164ykJAMADu3a6cT1dvwiXbBc6eKRbVLioFkJqyc4u8OBoavWl91J1bgQjjiMMYb/frIV5korXn99GuRyKdxur/hepVIJBEFARUU1OI47a5NR0jpR4EMIIReoQYO6nrWOx2KxIS5Oh9TUdiGfj0TQ4xGAHIsC2SYlTljlIY/RybxIq8nuxMjqr92pHZj5a3bOxWy2YtOmP7BvXw7++c9r0a9fKiQSXrxdZrc7a/ptqRvUZJS0PhT4EELIea6+BQjj43WQy6XizCw/rVaFTp3aQakMHXhEQoVTgmyTEofMSti9wROGeY6hm9aJfgYHOsa4zxnEMOabli6TSWqCHwZBqH+dIMYYLBY7du/+S/z+zHv0AyxbNhN6gwZWq6NmZpsAo1HbpCajpHWgwIcQQs5j9S1AuPiFr1FWVgWXywOFQga73VfL0r69HikpcS3SFd0jAMeqfNmdQlvoIMsg9yDN4EBvvQPqejqhhyIIAjwer9j6guPqXwmaMYaqKhtOn64MCGYyM/dj+vS3xO+fXh8Dl8vTrCaj5MJHxc11UHEzIeR8UV/hstGoRbXFjhkz3saj825Fv36pKCwsR/v2BnToEBfxcZU5zmR3nCEahEo4hh46J9IMDiSrz53dCcXl8sDj8UKhkIntMTwe38KEPM9BJpMA8N2+crnc+Pnng/UGMw3px0UufFTcTAghF7D6Cpc5joPFYofBoMFzz9+JNat3olevFPTokQKtVhWx8bi8wNEqJbJNShTbZSGPiVd4kGawo5feCaWkuYEFEzM9giCA43hx2r3HI9S0pJDB7fZgzuz38O679a/HQ/24SG1NCnzUajUee+wxjBo1CgkJCWITO79u3bqFZXCEENKW1M5MxMfrAgqXtVpVzfo7ckgkPHieQ1ycDn36dATARSToYQw47ZAiy6TEkUoF3CGyOzKeoafO10KivcrT6OyO0+lGSYkZ7dvrIZfL4F8lmuM4uFy+DJfvOAe8XgFKpTygUNlqdeD3349RBoc0WJMCn/feew8jRozAxx9/jKKiovPuf7gZM2Zg7ty5SEpKwoEDB/B///d/+OWXX6I9LEIICcBxHAYP7oZhw3qgR88UXHJJDyQlGSGXSyHhecTq1cjPd0MqlSA1NQESCQ9BYDUFvr6aF61WFfZ6HqeXw+FKX+1OqSN0dqe90o00gwM9Y52QNzG7wxhDbk4JrDYHHA43UlPbQSaTwul0IS/vNFQqOTp2bAe324PCwnJYLHao1Qqxe7zD4UJSkhHx8brmvF3SxjQp8Lnuuutw/fXXN7o/U0u4/fbb8eqrr2LmzJnYvn07pk2bhnXr1qFPnz4oKCiI9vAIIQQcx+G++8ZgzsM3o1OndpDJJGLm3G534eTJckgkHIxxWqSmJsDt9kAi4cVC5jPlmVzYgh7GgCK7FFkmFY5VKuAJ0SBUzgvoFetEmsGOBFXoXlkA4PX6Vkv21eGEeq0zWR25Qga7wwmPxwOr1QmlUkB1tR16fQwAwOXyZYQsFjsAX6sNP5VKLq5STUhDNam4OScnB+PGjcPhw4cjMKTm2blzJ/bu3YuZM2eK+w4ePIjVq1dj/vz55zyfipsJIZGUnt4fi5dMwcCBXcWi3bo8Hi9OnaqAXq+BRqOsabIpAGCQSiVhzfDYPRwOVfpqdyqcof8tnKx2I81gx0U6J2TnaG3tdLqRk1OM1NQEqNWKgOf8dwd8dTsMXq8As6kaXkEQZ6stWbwKZnM14uN1KC+3YOG/7xaLt+tKSYlDVlY+rhubcd7deSAtL6LFzU899RSeffZZTJ48GXa7vcmDDDeZTIYhQ4bghRdeCNi/ceNGXHbZZSHPkcvlUCjO/HJqtdqIjpEQ0nalp/fHsndmolOnBPA8V9MINHj1ZKlUgo4d4+HxeMV1a+rLnjQFY8BJmwzZJiX+qlIEdUIHAKVEQG+9A2l6B+KU9Wd3avN4BOTnl9aMVxoQ6NT+L2O+IKi8vApzH1mBsrKqemdbLX7hayxbNhMpKXGoqLAEzGyjtXhIUzQp8Hn44YfRrVs3lJSUIC8vD2534OJZQ4YMCcvgGis+Ph5SqRQlJSUB+0tKSpCYmBjynMcffxzPPPNMC4yOENKW+WdpGQxasUYnVNBT+3iZLLwTb60eDgdrGoSaXaGv3THG10Kim9YJ6TmyO3WVlVXCYrFBp/M1Q2YscMFBf6AHcJBIeFgsdnz22U9nDVzqrsVjNGppLR7SLE36rVq9enWYhxFedX+J/DMAQlm0aBGWLl0qPtZqtSgsLIzo+AghrVvddWP++CMXkyZdiYsvvggqla84t+bIiI9FYMCJahmyTSrkWOQQQrymWiqgj96BNL0dekX9LSTOxb+CtMfjredvLgtoPeHxNCyTlJm5H1u3ZtFaPCQsmhT4PPvss+EeR1iUlZXB4/EEZXcSEhKCskB+LpcLLldw915CCGmKuist8zwPiYSHQiGFRhO5dXbqsrh5HKjJ7ljcoW6TMaRq3OhnsKOL1gVJM2MwxhgMBg3Kyqpgsznhdnsgl0vrNELlwHEMAIPD4UZMjBKDBnVt0Bo7tBYPCZdm5VEHDx6M3r17gzGGgwcP4o8//gjTsJrG7XZjz549GD16dEBWavTo0VizZk30BkYIaXVCrQY8cmS/gJWWZTIZUlPb1crwRJbAgFyLHNkmJfKq5WAhsjsaqRd9axqE6uRNz+6IrykwCAIDY77Ca38tjtlsRUyMUsy4u92+7I5EwsHjEVBQUAatVkVT0UmLa1Lg065dO3z++ecYOXIkzGYzOI5DbGwsMjMzMWnSJJSVlZ37IhGydOlSfPzxx9i9ezd27NiB+++/H506dcKyZcuiNiZCSOtSX/8so1EDrU6NigoLZDIpkpMN4HnfQnyRbBha6TqT3bF6goMsDgxdtC70MziQqnGhnj6fjea7peXrq1VcbIJKJUd+3mkkJRuhVMggCIJYyOwPgKxWX3d0j8dLU9FJVDQp8Hn99deh0+nQt29fcUp779698eGHH+I///kP/v73v4d1kI3x5ZdfIi4uDk8//TSSkpKQnZ2NcePG4cSJE1EbEyGk9aivf9agQV1hMGjg8XgRF6cFx3HigoPSRjTnbCivABy3yJFtUuGENXRQpZN5kVaT3YmRNT6746/VqT2F3l9X4/X6Mj0OhxvFxSZYrXYkJRmxYMFnKC2tRLt2sXg64w507pwAk6laXHTQvw6Pfyr6vn05TfwOENI0TVrHx2w245prrsHu3bsD9l988cXYuHEjDAZDuMbX4mgdH0JIfTiOw7r1C9C/f2ecPFkWsIpwu3ax4m0bf81tJBqkVzglOGBS4qBZCbs3eNoVzzF00zrRz+BAx5imNQj1czjcNcGbAIvFBqNRi1OnTLBYbEGBjEolh0ajwq0T/i3W4viDRI1WFXIq+ozpb9GsLBI2EV3Hh+f5oCnsgK/Gpm7fLkIIibbmdOcO1T/LbnfhootSoFTKalYgRp06nvqnqTeFRwCOVflaSBTaQmd3DHIP0gwO9NY7oA5Thkki4WC3O2syOg7odGrExChQUmIKOtZo1AZlcGgqOjkfNSnjs3r1auj1etxxxx0oKioCACQnJ+O///0vTCYTJkyYEO5xthjK+BDSutRXj1P3g7e+YmX/uWq1AgqFDGq1QgyaPB4vBIFBpZKHvV8WAJQ5JMg2KXHIrIQzRINQCcdwkc6JNIMDKermZXfqYoyhoKAMpaWVAHwZnfh4HZxON2QyaaMyOM0JPAlpqIZ+fjcp8OnQoQPWrFmDtLQ0FBQUgDGGTp06ISsrCzfddNMFvQ4OBT6EtB711eMYjVpUW+yYXvNBHSo4Ki2tRGKiHlqtGkqlHDJZYKsI32wmIewzttwCcLRSiSyTEsX20A1C4xUepBns6KV3QtnEBqHnwhjD4cMng2pylixehUfn3XrOQJKQlhbRW10nT57EkCFDcM0116BXr17gOA4HDx7E5s2bmzxgQggJJ/9KyVqdGidPnplpare7UFhYjpSUOMx77DZwHIe3354RFBz179+lpqg39PV5ngPPhy/oKbFLkW1S4kilAq4Q2R0Zz9BD50CawYFEladJ2Z26LSTOtZ/jOKhU8oD2EJmZ+5GZuZ8yOOSC1aSMT2tGGR9CWofBg7th1TfzUV1th90evEipvxi36FQFOndpLwZHarUCiYlGGAwxER+j08vhcKWvdqfUETq7017pRprBgZ6xTsibmd3xd033/9n3z9byt5LwPecPgBjMZhuqq+2U0SEXhLBnfB588EG8++67cDqdePDBB8967Ouvv97wkRJCSJjUriXp3j0JcrlUbKNQm1qtgEwmhU6ngkqViNLSSmi1aiQmGhATo6i3a3o4MAYU2aXINqlwtFIBT4gGoXJeQK9YX+1OgsoTltd1uTzIzz8Nr1dAfLwOTGCw2Z2Ij9dBofAFXb7p6S6YTNVQKGRY+PwX2LnzCGV0SKvS4IxPTk4Ohg4dioqKCuTk1L/uAmMM3bp1C9f4WhxlfAi5MNWt0wEAvT4GJSVmcZE8f3CjVMpqblVx8HoFmCqqYTD6modGKuixezgcqlQi26REhTP0vzmTVG70M9pxkc4JWZiH4fEIyMkpgsfjFTNdqZ0TUFhYHjAt32ZzivU8143NoICHXDDCnvHp2rVryK8JISTazlbE3KFDfE3Wh0NqajtIJDw8Hi94XgK32wu5XIaE9vqIjIsx4KRNhmyTEn9VKeANkd1RSgT01juQpncgTtmwpp1NGQfAkJhogM3mFIuU3357hthiorraDoVChpSUOLGeh4Ie0ho16d8UTz31FFSq4GZ7SqUSTz31VLMHRQghDVW3iNlud9XMwpKiuNi33kznzu3RsWOcOANLJvP9m08i4SOyyKDVw+H3UhU+/MuAVXl6HKlUBgU9HWJcuK5DFf7ZoxwjEq0RC3oAX0sJr1dATIwSLpcHi1/4Glu2/Inp099CVlY+NBoVkpKM0GhUyMrKp4UFSavWpOJmj8eDpKQklJaWBuw3Go04ffo0pNJm9T6NKrrVRciFpXYRs1QqFW9l+XtDAQiaih4JAgNOWGXINqmQUyWHEKJBqFoqoI/egTS9HXpF8xuENgRj/qJm3+s9/dQnePHFb8TnaY0d0lpEdDp77T8otQ0YMAAVFRVNuSQhhDRJfLwOcrkUMpkMnTrFQyKRQBB8s5d4noNMFtnO6Bb3mQahFneo12JI1bjRz2BHF60LksjGX4GvzBhKSytRWWmDRMJDKpVg8+Y/g47xt5ggpC1oVOBTUVEBxhgYYzh69GhA8CORSKDRaKgLOiGkRZWVVcHl8qBDhzixqDnSWWeBAbkWObJNSuRVy8FCZHc0Ui/61jQI1clbJrsjjk9g8Hp9ndPLyy0BBcvUFJS0dY366/B///d/4DgOK1asQEZGBiorK8XnXC4X8vLysHPnzrAPkhBC6vLfomnXLhZerwCVKnQPq3CqdJ3J7lg9wdkdDgxdtC6kGRzorHGBb9HsDlBVZRPbZ3AcB7vdBcYYFSwTUkujAp+PPvoIAJCbm4tff/0VHk941pcghJCG8Ac7o0YNwM23XIrkZCP0+hhoNMGTLcLFKwDHLXJkm1Q4YZUBIbI7OpkXaQYH+ugd0MjCn90RBAa32wOr1QmZTIKYGCX4OlGVIAjgOA4mkwUGgwYcx4tT16kpKCFnNCkf/NNPP4lfK5VKyGSBK45SUTAhJNz86/QMGdIder1aXHG4bgAQLianr0HoQbMSdm/wBFieY+im9S0y2CkmvA1CBUHAxx9lwmZzwGJxYNWqX7F373EMHNgF8fE6dO2ahKeengiNRlkzi00GpVIBjUYJtVqOP//MxXvLNyIvr4QKlgmpo0mBj0qlwpIlS3D77bcjLi4u+KIX8KwuQsj55+qrB+C/nz6CuDgNJJIzt5jCPVHLIwDHqnwtJAptoW+dGeQepBkc6K13QC0NfzDh8XjBcRw2bfoDn366LeC5vXuPg+M4rFt/CziOw5EjZxpC+xchNBg0qKioxvLlGyjYISSEJkUoL774ItLT0zFz5kx89NFHeOCBB5CSkoJp06bhscceC/cYCSFtFMdxuP/+sXjl1X+KbRUioczhy+4cqlTCGSK7I+EYLtL5sjsp6vBmd/zOBCm+1aTLy4Mz5xzH4Y47rsKAAV1QXW0PeM7fRd3t9qBnzxQMGtSVZmsREkKTAp/x48fj7rvvxrZt27BixQr8/PPPOH78OPLz8/GPf/wDn376abjHSQi5gDVmrZjadTyTp4xCz54p4Pnwt5FwC8DRSiWyTEoU20MHVXEKD/oZ7Oild0LZzAahZ+P7VviWVRMEAWazFaWllQHH+G/1DRzQBXHxOuj1MYiP16G42ASL5UwQ5HS6YTRqER+vi9h4CbmQNSnwMRqNyM3NBQBUVVXBaDTi+PHj+OWXX/D222+HdYCEkAtb7R5aarUCgsBQUmLCp//dhk2b/ggIgvzHDhjQBXFx2oj0zSqxS5FtUuJIpQIuIfj6Uo6hZ6wDaQYHElWeiGR36uI4X/Bjt7vgcLjx55+5AdPOa7fksFTbEauPAWMMKpUCqakJyM8/LQY/CoUMLpdH7E9GCAnUpMAnJycHnTt3xokTJ3Dw4EHcfvvt+P333zF+/HiYzeYwD5EQcqHyf2DH1SwyKJfLIJXySEiIxb8Xdca8x27D778fw+IXvgYALHtnJhIS9NBolGHN8ji9HA5X+mp3Sh2hszsJSjfSDA70jHVCEaHsjj/Aq12U7fUKcDrdqKy0guO4oGnndVtyAEB8fCxUKjncbg9kMt9q1f7Ax2jU0no9hJxFkwKflStXYsCAAfjpp5+waNEirF27Fg8++CCkUinmzJkT7jESQi4wHMdh8OBuWPLiPUhor4dCIYNUKqn5sOfEVZV1OjWGDOmOZe88AJlMgg4d4sOW5WEMKLJLkW1S4WilAp4QDULlvIBesb7anQRV5Jfn8C0AC5SXV6G8zAKpTAKt1jcV3+324siRwqBp54MGdUXPnikoLz+TwSkuNiE1tR1kMikEQYBSKYfBoIFaraD1egg5hyYFPq+++qr49datW9GrVy8MHToUx48fx/79tE4EIW1R3TV2OndOQLt2sQHTzTnOF/QAviCA5zloNErExqrDFvA4PBwOVSqRbVKi3Bn6T1ySyo00gx09Yp2Qhf9uGgBf4FX7NpkgMBw8cAKffLIVmzf/KWZkzlX75G/J4esw72Ox2JCfXyr2JZNKJdBoVPjzz1xar4eQcwjLvPOCggIUFBSE41KEkAtQ3docwDe7iOMQ1ByU5329/vz7/W0mmoMx4KRNhmyTEn9VKYI6oQOAUiKgd03tTiQ7oQeOyxfECALD7t3HcMXl88RmoX7nmnnlb8mhUMhgt7vE/RaLDRaLDQaDBlqNCg/PeQ+fffYTZXoIOYcm/8W5+OKLMXLkSCQkJATdi3/44YebPTBCyIWhduGtWq0AYwxutxdyubTeOp1wdUq3eTgcNPuyO2ZX6D9nHWJ8LSS6a52QRii7UxdjgMfjAcfx4Djfrakn5n8cFPQ0xL59OThypBD9+qWisLA86Hm1WoE//syloIeQBmpS4PP444/j+eefx5EjR1BSUhLwy0a/eIS0DbXreAxGLUpPV8Jo1EIQWMhMT7gwBuRbZcg2qZBTJYcQooWEWiKgt96X3TEoWia74xvbmUwWz/Nwu704cCAf8x79oMm3nxhjWPzC11i2bCZSUuJQUWGB0+mGQiGD0ailmh5CGsm3cEQjFRcXY968efjwww8jMKTo0mq1qKqqgk6no9YbhNTDf2urX79UtGsXK9br8DwfVNsSLtVuX4PQbLMSFndwg1CAIVXjq93pqnEhAjPhz8rrFeD1CigpMWP1tztw9Ngp7Np5BHv3Hg9LUFJ7WQC5XAqXyxOyGJqQtqqhn99NyvgIgoDt27c3eXCEkPObP5szbFgPcByHykobYmNVAMchOSkOd9+dDoVSLq4eLJHwYqYjrD2rGJBXLUdWhRJ51XKwENkdjdSLvjUNQmPl4W8Q2hC+2VoMWVl5zcrunE1m5n5s3ZrV4IUgCSGhNSnjM3fuXCQnJ2P27NkRGFJ0UcaHtHXp6f2xeMkU9O3bCXK5LGBWVu1sjsfjhdvthVIpC/ttrUqXL7tzwKyE1ROc3eHA0EXrq93prHEhQn1KG8Tp9ODAgXwsf3cD9cciJIoa+vndpMCH4zisXbsWPXr0wMGDB+F2uwOev/XWWxs94PMFBT6kNWlMqwjAF/R8+OH/ITHJCI7jItb5PBSvAORY5MgyqXDCKgNCZHd0Mi/SarI7Glnkszv+TI6/SNv3vePAcb7Md0VFNZ568hMKeAg5D0T0Vtfrr7+O9PR0ZGZmory8nH7hCTkP1VcTsmTxKpjN1UHBkH+F4HYJegBosaDH5KxpEGpWwhaiQSjPMXTTupBmsKNTTGQahIbidnvhcLjAGINSKYNM5stsMQa4XJ6I3tYihEROkzI+VVVVmDRpEv73v/9FYEjRRRkf0hrUnmJeXl4lzgLyLXgnR3W1AxIJD8ErICenGE899Qnat9fjtf9Mg06nCqjZiQSPAPxVpUCWSYlCmzzkMXq5R8zuqKUt+48rQRDw11/F8Hg80GhUePKJj6HXq9GhYzucLCjDzjAWLRNCwiOiGZ+KigocP372RbcIIdFRt7eTWq2ARqOCXC5FTIxS7Jnln4kV306H9RuehcPhglIpj2imp8xRk92pVMIZIrsj4Rgu0vlaSKSoWy67U5vHIyAnpwgWix08z8Fo1KK0tBL//e/Wlh8MISTsmhT4PPPMM1iwYAHuuece2O32cI+pyXJzc9G5c+eAfS+88AIef/zx6AyIkCgYPLgb+vVLhVQqQd++nWpaQXABWRyplAsoVOY41Cw+GP7xuAXgaKUC2SYViuyhG4TGKXzZnd6xDihbOLtTm93uQkFBKXU6J6QVa1LgM2vWLHTr1g0lJSXIy8sLKm4eMmRIWAbXFE899RSWL18uPq6uro7aWAhpaenp/fH2splITDQE7PctKsgFZFBCZVPCmWE5bZciy6TEkUoFXEJwdkfKMfSsaSGRqPK0aHaHMV9LjYqKahgMMZDJpLDbnTh8+GTAcdTpnJDWp0mBz+rVq8M8jPCxWCwoKSmJ9jAIaXG1Z2TV1VKFyk4vhyOVCmSblDjtCJ3dSVC6kWZwoGesEwpJdLI7brcH+fmlcLvdkMul0Go5OJ1uqFRyWhWZkFauScXN56vc3FwoFArI5XIUFBTgq6++wosvvhiUkapNLpdDoVCIj7VaLQoLC6m4mVxQeJ7Hz78sxtChF0HaUg2pajAGFNmlyDYpcbRSCU+IBqFyXkDPWCf6GRxIUHladHx12WwOVFc7IQiCONNtw4a9GDNmMK2KTMgFLKLFzeer1157DXv37oXJZMIll1yCRYsWoUuXLrjvvvvqPefxxx/HM88803KDJKSZ6q7No9dr8Pzzd+LiSy6qtyloJDg8HA5V+hqEljtD/ylJUvlaSPSIdULWQkPzeLzwegXIZNKATJfT6YbV6sB1Y5+BIAhB0/lffWUNrYpMSBvQ4IxPeXk5evTogfLyclRUVJz1D0JcXFy4xoeMjIxzBiZDhw7Fnj17gvZPmDABq1atQlxcHCoqKkKeSxkfciGpuzYPz3PQaFSQySSQySL/7xjGgEKbDNkmJY5VKeANkd1RSAT0rqndiVe2XINQ//gEQYDT6cvy8jwHjvMVdldV2bFnz1+4bmwGBTSEtEJhz/jMnj1bvNDs2bNb7A/HG2+8gc8///ysx+Tl5YXcv3PnTgBA9+7d8dtvv4U8xuVyweVyNWuMhLQE/9o8BqMW1dV2mM1WdO6cENRWIhJsHg4HzUocMClhcoX+s9FB7Wsh0V3nRAvfbQPgW3DQt6Iyg0wmhdfrhSAwSCS+4Ke6mup1CCGtrManruuvvx4//PADOnXqhIKCggadQwsYkvMRx3HY9dvL6NevMwBfU1BJhNuPMwacsMqQbVLhuEUOIUR2Ry0R0Fvvy+4YFC2b3fHzBzKMAdXVDhQXm2oWapSJbTfMZismTVyCLVv+jMoYCSGRF9EaH4/Hg6SkJJSWlgbsNxqNOH36NKTSli8dGj58OIYPH47MzExUVlbi4osvxiuvvII1a9Y0OOghJBIa2y8r1PnPPfcPDBzYteYxIlrLU+3mcaAmu1PlDm4QCjB0inGjn9GOrhoXIhx/nZPvW8kgCAwejxcejwfHj59CbGwMDAYNqqvtmHrPfyjoIYQAaGLgU99S9gqFImq3jZxOJyZOnIiMjAwoFArk5+dj+fLlWLJkSVTGQwhQf7+shswW4jgO9903BvfdPwb9+3eOaIZHYEBetRzZJiVyLXKwEA1CY6Re9NU70NfgQKw88g1CG8NksuKFRV/j2jGD0LNnCoxGLVwuD3bv/otmZhFCAjTqVteDDz4IAHjllVfw1FNPBSwOKJFIcNVVV6Fz584YPHhw2AfaUuhWFwmX2v2yrFYHBMHXIkKtVqDaYseMGW+HbBbqP3fxkikYMKBLRPtmVbp4HDApccCshNUTnN3hwNBF66vd6axxoQWbtZ+Vr2u6L/tVUmLGXXcuxZYtfzY7u0YIuXA19PO7UYFPTo5v9dLU1FScPHkSXu+Ze/oulwt5eXl4+umn6y0kvhBQ4EPCgeM4rFu/AEOHdodEwkOplNcELwxut6/o1j+9um4miOM4rFg5C8nJceB5PuwrGnsZkGORI9ukQn61DAiR3dHJvOhrcKCv3gGNLPrZHf/tLKfTI65TxHFUu0MIOSMigY/fli1bMGHCBJjN5mYM8fxEgQ8Jh8GDu2Ht/zKg12vA8xy8XgFSqUScXu1ntTpQbbHDZnPBYNSA4zgoFDIoFLKwBzwmZ02DULMSthANQnkwdNO5kGawo1NMdBqEno3d7sLhwycRG6um2h1CSJCIFjdfffXVAY95nke/fv2Qn5/fKoMhQhqrXbtY6PUx4HkOHo8AhUIKjuOCbrvExCgRE6OM2Dg8AvBXla+FxEmbPOQxermvQWgfvQPqKDYIDYUxwOsV4PV6YbHYkZRkoNodQkizNCnweeWVV5CVlYUVK1aA53n89NNPuPTSS2Gz2XDDDTdg27Zt4R4nIeeFhtaQxMVpIZFIIAhCQPYmUrU6dZU7JMgyKXG4UglHiOyOhGPornMizeBAB3V0szterwDGBPC8JKCRKmMMJlM19uw5jiWLV9VbD0UIIY3RpMDnb3/7Gz755BMAwPjx49G5c2f06tULd999NxYuXIgrrrgirIMk5HxwthlaW7dmBQREFRXV4kJ6LcUtAEcrFcg2qVBkD90gNE7hy+70jnVAGeXsDmO+OieXywOlUg7GfLVEbrcHVqsDYMCcOe/hs89+oiCHEBI2TfqrHBcXh+LiYgDAuHHj8NVXX+HYsWN4//33MWvWrLAOkJCWxHEcBg/uhmHDegAch107j2Dv3uMYObKfOEOrvLxK7OA9eHA3fP7FoygtrYTBoIFCIYMgMFgsNkgkLZNGOV3TIPRwpQIuITi7I+UYetS0kEhSeaJeu8MYq8l8cSgoKENVlQ1qtQJSqQQejxc2mxM8zyEpyYiysioKegghYdWkwKekpAR9+vRBUVERxo4di5kzZwIA1Gp1wEwvQi4k/inkfft2EjM1brcXBw7kAwC0OjVOniwDAMTHx6JdOx3kcimkUgni43UB14qL00Z0rE4vhyOVvtqd047Q2Z12Sjf6GRzoGeuEQhL94KF2/OKbjs7EtYlsNmfAsQqFDC6XB2VlVS05REJIG9CkwGflypX48ssvUVRUBMYYfvzxRwDAsGHDcPjw4bAOkJCWkJ7eHx9++H9ITDKCMQaXy5cZkUol4orJJ0+WQatVoVOndlAqQxcKRxJjQHFNdudIpRKeEC0k5LyAnrG+2p32Kk+Lj7E+Xq8At9sjNgy12VxQKqUwGDQwmaqDjjcatcjKyse+fTlRGC0hpDVrUuCzYMECZGdno2PHjvjqq6/E1Zq9Xi8WLVoU1gESEmkcx2HeY7ehXYJeDHoAX6Dh8XhruqDzaN9eX1Oo3LL3ihweDocqlcg2KVHuDP0rm6jyZXcu0jkgD9VlIooY87WSAHwzQL1eASaTBSqVHE6nGykpcaiosIi3D41GLSwWaihKCImMRq2Bv3btWuh0vpT+qlWroFarA1Zv/uGHH/DCCy+Ed4SERNigQV3Rr18qOI6D2+37gJZIeHE9HX9frDOLEEYeY8BJqwzrT2qx/GgcthVrgoIehUTAQKMNd3arwKSuZvQ1RC/oOVeA4lvDiIfd7kJ+fikUChn+/DMPU+/5D7Ky8qHRqJCUZIRGo0JWVj5mTH+LpqoTQiKiURmfMWPGQKFQiI/nzZuHzz77DJWVlb6LSaXo2bNneEdISITFx+sgl5/p5M1xPGQyibjuzpli3MizeTgcrGkQanKF/vVMUbvQz+BAd50T0ig3CPXzt4/wfc1QUWGB1epEYqIBUqkEjAkoKTHBYrEHZHQyM/cjM3M/tZkghLSYRgU+df/4t3TKn5BI6NK1PTQaJaRSvqYv1pnnWuL/ccaAE1YZsk1KHLcoIISo3VFJBPTR+2ZmGRTn3wQCvqaJF2MMx46dgsViBwA4HC4kJhoQE6NE+/YGMAZkZeUHLD7IGMPevcejNnZCSNvScouMEBIF51pwcPacm7Fw4V2Qy6U1x7fc2KrdPA6afbU7Ve7Q96g6xfhaSHTTuhDB5uz18n+v/Fkvxph4608QBHi9vsarTqcbcrkUJ0+Wi0EPAFgsdlgsdhgMGmg0Kjwx/yNal4cQElWNCnz8af+6+wg533Ach/vuG4P77h+D5GSjuP/UqQpsWL8XP/98AJ07t8fChXdBoQg9HTwSBAbkVcuRbVIi1yIHC9EgNEbqRV+9A30NDsTKW75BqCAIYAyornaAAxCjUcJmc6CoyASvV4BWq4ZeHwO5XCpmxE6dqoDRqIXJFLo/TmWlFWq1gtblIYREXaNvdX3wwQdwOn1rbiiVSixbtgxWqxUAAup/CIkW/3o8AwZ0EbMTPgyJiQYMHtwNj7gngOc5cR2ZSKty8ThQU7tT7QnO7nBg6KxxIc3gQBetC3wU7yKfOFEGpVKGhQu/BAA88cTtqKiwwG73zd602ZwoKTFBrVZArVZAoZDh2QWf4fmaINJ/XG20Lg8h5HzRqMDnww8/DHjsb1tR20cffdS8ERFSj4b0yUpP749l78xEp04J4HlOrD2puYL4lUwW+elPXgbkWuTIMqmQXy0LeH0/rcyLNL0DfQwOaGUtn92py+MRYLM5wPMcdu44jH37cnDjjcPQr18qCgvLA4612ZwwGDTIysrHZ5/9hLvuvjrkcQCty0MIOX80KvCZOnVqpMZByFmdrU+Wv0jWvx6PwaCFRMLXyfa0HLOTR7ZZhYNmJWye4DHwYOiq89XupMZEt0FobYwxFBdXBAQpjDEsfuFrLFs286zr7QiC0KDj6DYXISTaOAD0l6gWrVaLqqoq6HQ6WCyh6xVIy0pP7x/QJ4vneahUcmg0KphMFkyf5lvzZfDgblj1zXyAAZ1S27XoGD0C8JfF10LipDX0qs56uQd9a7I7MVFoEOoPOgSB1UzbPxNxeTwCyssrwZivILnuOjoNCTwbcxwhhIRbQz+/KfCpgwKf8wvHcVi3fgH69++MykobEhMNUCpltaZPA/v352HG9Ddx7bWDMHvOzXC53EhKMp7jyuFR7pAg26TEoUolHN7g7I6EY+iu87WQ6KCOXnbHNzHBV7j855+5ePyxj5B+dT/87W9XQK+PgcfjPWeQ0pBbjY05jhBCwokCnyaiwOf84s/iCAJDUpJvMTyO42oCCE4MJKxWB9xuL3Q6NQRBgFQauRoetwAcq1Ig26TCKVvoGWFGhQdpBgd6xzqgikJ2BzjTFJQxAWazFXl5p7H83Q1YvnyDGIhQkEIIaS0a+vlN6/iQ85pvVWUp1GplTdsDDv54vXb2JCZGKX7N85EJek7bJcg2qXC4UgGXEJzdkXIMPWKdSDPYkaTyRL12p7y8Csf/KsK33+7A5s1/hgxqaPFAQkhbQ4EPiajmZhTKyy2QyaSIiVHUqUtpmajC5eVwpNJXu1PiCJ3daad0I83gQK9YJxSS8yNb4vF4MWf2e/j0022UwSGEkFoo8CER09BC1/qCI/96PEajpkXbozAGFNulyDYpcbRKCbcQ/NpyXkDPWF/tTnuVp8XG1hCMMVRV2VFaWklBDyGE1EGBD4mIujOx/FObBw/uho8+noPnnvscy9/dgJEj+4UMjjZs2IvZ/3cjkpKNLRb0OLwcDpt9tTtlztC/GokqX3anhy56ndDPhjGG06fNcDppsUBCCAmFipvroOLm5qs9E+vkyTIAgFarFmdkSaUSuFxuHDhwQuzeXTs4Mhq10OnUkMmkkEp51C5iDjfGgFM2GbJMShyrUsAbokGoghfQu6ZBaLzy/GsQCvimqNvtThQWVkCnUyErKx/Xjc2gjA8hpM2g4mYSNYMGdUXPnikoL6+CWq2AVqtGQkIseJ6Dx+OF0+mbddW/f2fwPI/jx4tht7tgNGogl/umqvumrEduAUK7hxMbhJpcoX8NUtS+FhIX6ZyQtvBaiL6AhYPX64VUKgFjDB6PFxKJRMyA+YNBl8uDnJwSCIKXFgskhJBzoMCHhF18vA4ajRIGQwwUCjlkMt+HtdcrAODAmACOAziOB2MMqantIJVKIt43izGgwOrL7hy3KCCEyO6oJAL61DQINSqil91xOj1gjEGhkMHt9tcQcXC53AAAmUwKxhhcLg+qqx3Q69VwuTzIysqnxQIJIeQsKPAhYde5S3vodGoAqAl2fBkMnuehUHDi136RXHMHAKzuMw1CK92hX6tTjK+FRDetCy3Ut/SsqqvtEAQGxhgcDhcsFjsMBi1UKhkkEom4EOFj8z6E2VxN6/AQQkgDUeBDworjOEyYcBkEgQVMPWeMgeNYzeKDnNg2IVIEBuRXy5FlUiLXIgcLMf09RupF35rsTqw8+g1C/axWBybf/QpOnzZDr9fg0Xm3omfPFNhsDthsDpw6VYHlyzdg+bsbKMghhJBGosCHhJW/vufUqQokJxtrsjms5tbWmeAjUkFPletMdqfaE5zd4cDQWeOr3emidSGCsVeTeL0CrFYHTp82iwsLZmbup9WVCSEkTCjwIc3C8zwmTboSqakJyM8/jYqKamg0KnGVZZ7nI76CsZcBuRY5sk1K5FXLEWpxQ63sTHZHK2vZ7I4gCOcs1GaMwW53obCwHFqtCvHxuoDnaHVlQggJDwp8SJPNnnMzHn/8b9DrY8BxnFiPolLJwRhq9dSKDLOLR7ZJhYNmBWwhsjs8GLrqfLU7nWLcUcvueDwCTKZKGI1acBwn1jQxxuB2e2E2W1FeXgWbzQmVSg6Xi9bgIYSQSLlgAp/58+fj+uuvx8CBA+FyuWAwGIKO6dixI958801cffXVsNvt+PTTT/HII4/A7XZHYcQXprO1mKj93Jgxg/GvB28Az/PweLw1U635gJ5ZkeARgOMWXwuJAqs85DF6uQd99Q70MTgQE6UGoX4ejxd5eSXgOA7x8bGwWh2Qy6WQySRwu33T02Nj1TCbrQAAo1GLrKx87NuXE9VxE0JIa3XBBD5yuRxfffUVduzYgXvvvTfoeZ7nsXbtWpSWluKKK65AXFwcPvzwQ3Ach1mzZkVhxBees7WYACA+p9EoYTRqxfNkMt9UdEEQaoqYw59aqXBKkGVS4pBZCYc3+LaRhGPorvO1kOigdke9QSjgy/Tk5BTDYrGjT5+O4Dhf41C324vU1ARIJBJ4PL51epKTjaiuVtEaPIQQEmEX3MrNkydPxquvvhqU8Rk7dix++OEHdOzYEUVFRQCAiRMn4oMPPkBCQkKDV2Fuqys319diwmjUwuvxAjW3aBwOF1JS4iI+BR0A3AJwrMrXQuKULXSDUKPCgzS9A731DqiikN2pb3aaw+HCiROlAICUlDjExCjFW1sOhwtVVTbodGoolXKx+eru34/hiSc+pjV4CCGkCdrcys2XXnopsrOzxaAHADZs2AClUokhQ4Zg69atIc+Ty+VQKBTiY61WG/K41ozjOMx77DZodWqxxQQAsdi2f/8u4Djg2LEidOoUH/Gg57RdgmyTCocrFXAJwdkdKcdwUawT/Qx2JKk8Ucvu+NYj8r24IDBUVdkgkXBwOj04fdoMmUyK1NQEyGS+75fL5YEgMKhUCigUMuTnn4bXK0Amk0Kvj8GCBZ9R0EMIIRHWagKfxMRElJSUBOwzm81wOp1ITEys97zHH38czzzzTIRHd36r3WKiNq1WjeRko/jB3aNHcsRWV3Z5ORypVCDbrESJPXR2p53S1yC0V6wTCkn0EpWMMZw6VQ6n04P4eB2cTjcynv4U7767PqDpamKiHhIJD5vNCYVCBsZYTdbHA5lMisREA44dOwWVisFmc6K0tDJq74kQQtqKqK5Rm5GRIX4Y1LcNGTKkwdcLVRfhn21Un0WLFkGn04lbSkpKk97LhSw+Xge5XAqn011rXyy6dEkQZ2gBCHvQwxhQbJNiU6EGy4/GYXORNijokfEC0gx23NHVhH90M2OA0RG1oMf//6TJZIVEIoFUKsGuXUdxx6QX8c4768AYQ2bmflw3NgNPzP8IlWYb8vNP4/DhQthsLkgkZzJlHo8XSqUcarUCRqMWR44UUkEzIYS0gKhmfN544w18/vnnZz0mLy+vQdcqLi7GsGHDAvbp9XrI5fKgTFBtLpcLLperQa/RWpWVVcHl8kCh8HVOT0w0QKtV1QSNCPutJIeXw2Gzr3anzBn6f8FElS+700PngDzy5UTn5PUKsFjs+P33Y3hi/keIi9PWu5ggYwxlZVXwCgIqK60AGIqLTUhNbQeZTAqv1yvWBrVvb0B5eRUVNBNCSAuJauBTXl6O8vLysFxrx44deOKJJ5CYmIji4mIAwLXXXguHw4E9e/aE5TVaqz/+yMWpUxXo27cjlEo5pNLgDuDNxRhwyiZDtkmJo1UKeEM0CFXwAnrpfbU78croNQity+32QCKRwGZ1YvELX2PPnr/OeU7tYNJud8FisSE/vxSJiQYolTJIpb6C5qNHC/EkFTQTQkiLuWBqfDp27Aij0YhOnTpBIpFgwIABAIC//voLVqsVGzduxMGDB/Hxxx9j7ty5MBqNeOmll7B8+fI2NTursfxT2Hv16gCtVh3269s9HA6alcg2KWFyhf7fLUXtayFxkc4JaZQbhHq9AhgTalac5sTN5XLjuec/b3CAsm9fDo4cKUS/fqkoLPQF9xaLDRaLDWq1AomJBhw9Uogrr5gHQTh/+oQRQkhrd8FMZ1+5ciWmTJkStH/kyJHYtm0bAF9w9NZbbwUtYNiYW1ltaTq7fwp7XLwOWq0qbLO1GAMKrL7szl8WBYQQ2R2VRECfmhYSRkV0sjuMoWbtIX/9kq9thJ9CIYXT6YHN5sS+fTm4bmxGo25H+b+/Gq0KFRWWgCUCLBY7Zkx/izI9hBASJg39/L5gAp+W0hYCH47jMHhwNyx75wGkpiZALpNCq1M1+7pWN48DZgUOmFSodIcOojrF+FpIdNO6EKEJYmflz64IAoPXK8BsrkZ8fCw4DvB6mXiMb0FGhupqO8rLLU0OUs62KCQFPYQQEj4U+DRRaw18/O0mRo8egKn3Xov27Q2IiVHWfMhLmlzLIzAgv9rXIDTHIgcL0SA0RupFH70DaQYHYuUtd1uHMSYWETMGVFZawRiDWq0AwMFmc6K62i5OI+/YsR30+hhIJDy8Xl8PrT//zGt2kHK2NiCEEELCo80tYEjq5886DB/eE1ptYGaH55t2e6vKxeOAWYmDZiUsIbI7HBg6a3y1O120rog3CBUEBqfTjcpKK8xmK/T6GMhkUgheX9uIp576BCZTNeLjdSgv9/1C1J6ZBfjWM2rXLhZxcVqUl1tQWloZliCFuqsTQsj5gwKfVqBuRuGPP3IxcGAXxMfr0Llzezz88M1ITomryXQ0nZcBuRZfdievWg6EyO5oZV70rand0cpaJrvj8Xhhszlx8mQZNBoV7r5rKRhjjc6wUHBCCCGtHwU+Fzh/NmfQoK5QKGQ1NSy+RRu9XgExMUpIJLy4+nJTmF08sk0qHDQrYPMEX4cHQxetC/0MdnTSuCOe3amLMd86OU6nG0ajFnFxWmzcuK9lB0EIIeSCQIHPee5s9SHp6f3x+edzYYzThWyU6fUKNfsbH4l4BOC4RYFskxIFVnnIY2LlXqTp7eijdyBG1vI1K75vA8OpUxWwWOxQqeRwuTwoK6s616mEEELaKAp8zlMcx+G++8fgvvvGIDnZCAABM4K2bs3CW2/PQHy7WAC+rAdXp0K5KS0mKpwSZJt8tTsOb/D5Eo6hm9aJfgYHOsS4I9IglDHA4/HAbK5GbKwGMpkEbrcXxUUmxLfTQaGQwePxQirlYbU6xUDHaNQiKyufWj8QQgipFwU+LaQxM3vS0/tj8ZIpGDCgC3jeN8PIbnfD4bCgX79ULFs2E6+8shrduycBgDhzqancAnCsytdC4pQtdINQo8KDNL0DvfUOqKSRy+4Igq8fltXqhNfLYDZbodEo4XC4UG21w+nyiK0fPB4vSkrMUKnk4to41PqBEELI2dB09joiMZ29MWu5+Be965SaAJ7n4HJ5wPMcJBIJBEFAfn4pdDoVHA4XOndu36ygp9QhQbZJhUNmBVxCcHZHyjFcpHMizWBHstoTkeyOnyAwFJ2qQJXFBoNBg7y803h2wWcoLa2EXq/Bo/NuFb9/PM+LU84FgdHaOIQQQmg6+/nCH8hodWqUl1eJq/f6MzfTay2Mx3Ec5j12GwxGLQDA7fataCwIDILggUwmRWKiASdPlqJTp3Y15zRuPC4vhyOVCmSblUGd0P3aKX0NQnvGOqEMUyd0/604QWA1va98KyVzHA+vV0Be3mlYLDbx2ORkI0pLK8WZVpmZ++uduUZr4xBCCGkoCnwiyB/IaHVqVFRYoNGooNGoYLU6UFhYjpSUOMx77DZs3ZoFxhgGDeqKnj1TUF1th06nDvog93q9UCpl4Hke/oLlunU9oTAGlDikyK5Q4kiVAu4Q2R0ZL6BnrBNpBgfaK8Ob3WGMwe32QiqVwOVyw+FwQ6dTw+PxwuFworjYJAY9AMTZWfHxuoBr1J1uTtPPCSGENBYFPhE0aFBXDBjQGVqtGgkJsWKQIgi+nlBlZZXo2TMFgwZ1xd69xxEfr4NcLoXZbBUzJLWDH0FgkEo5JCbqoVKFnmlVm8PL4YhZgSyzCmWO0D/q9io3+hkc6KFzQh6m7E5dHMdBKpXAZKrGHZNeRPv2sVi69J+wVNthMlUHHa9QyGh2FiGEkIigwCeCRo0agPh4HXieB2O+rAVjAM9zUKsVSE42wmJxiJmNsrIquFweCIIAh8MFlUoBt9sDAOB5HnK5BDzPQ6/X1PuajAGnbFJkm1Q4VqWAJ0SDUDkvoLfeV7vTTtm8BqG1W0L4gjXAv45Q7cCt9HQl7rzzZWzZ8ic4jsPdk0ehX7/UkIEPzc4ihBASKRT4RAjHcbhlwqUBQY+fIDAxC6JWK8QWCvv25eDIkUIMGtQVVVU2KJVyX9sFQYBcLj3rbS27h8MhsxLZZiUqnKF/rClqF/oaHLhI54SsmQ1C/W/HZLLAaNSB43w1SYwxSKUSSCS8uM6OyWQVgx7fuQyLX/gay5bNREpKXMjO5TQ7ixBCSCRQ4BMhgwZ1RWpqQs2Hd3DAwpgvU1I7lhk5sh+MRg30+hjwvLbmXAapVCYeV3u9HsaAk1YZskxKHLco4A2R3VFJBPSuaRBqVDQtu+PxeGuKkc+M1x/oWK1OlJcXIrVTAqQyibhitMvlgdvthdlcjXun/kcMevwyM/dj+vS3xNluRqMWLpcHWVn5NDuLEEJIxFDgEyHx8bqaFhIMPI+geh1/8OJyuREXpw2Y/XXyZDmMRi1UKpnYOZ0xwO323QbzSBQ4aFbigFmFSlfoVhQdY3wNQrtpnZCeJbsTauHD2jweLxgDHA7fYoU8z0MqlaC8vAovv/QtNm/+E/v25SA9vT+ee/5OdO2aCJ73dT4/1xTzzMz92Lo1izqXE0IIaTEU+ERIWZlv6rogCPB6fason2kfwWp6agFWqxPt2unwf7NvRruEWJw+XQmPx4tjxyxQq+WIj9ehXbtYeAUBOVUyHKmOQb5VCSFEFkkt9aKv3om+Bjv08oY3CK0v+PF4vCgsLIfBoIFSKQ8I1p5++r9495314rFbtvyJzMv3NzqIoc7lhBBCWhItYFhHuBYw5DgO69YvwIgRaeA4Dm63p+bWli/z48/k2GxOMAbodOqazI7vVpHN5kJxsQlMroJJnYJskxLVIRqEAgydNb7sThetC5IGTkNnLHgNIP+qyf4Ap7ragWPHCgEAarUCUqkEBoMG+/bl4LqxGZSZIYQQct6gBQyjzF/A26dPRyQmGiCX+1osCAITVx/mOECrVcFfO+MPOjiJBCVMh2xZAk5YFYA1OJrRyrzoq3egr8EBrazh2Z3a46tds+Pf53C44PUKUKkU8Hi8UKnkcDrdYIxBq1XBZKqmwmNCCCEXLMr41BHulhX+vlt9+6ZCJpP4AhvOf7vL9633B0GVLh7ZJiUOmJWwhcjucGDoqvVld1I1LoTqVHGumh0AEAQBHOd7TY9HqBkDB6/Xi6IiE44cKcSGDXsxZszgBrXZIIQQQqKtoZ/fFPjUEYleXRzHYfDgbrjr7nRMnToaKpVCXPvGy4DjVb4WEgXW0IsS6qQe9NXbkWZ0IkZ27poZ/2vW3e8vkPZ6BTGYyc8vhc3mgMGggUajwsNz3sNnn/0kBlBUeEwIIeRCQLe6ziOMMeh0atx803AolXJ4vV6YXAocMClxqFIJuzd42pWEY+im9bWQ6BjjxvbtB6C+vG+t6fFnblXVzvL4a4h8bSI8Yi2RyVSN2NgYMMYgk0nh9Qo4ebIcNpsDAFBZaYVarUBZWZUY3FDhMSGEkNaGAp8W4O/ZpdSoccisQLZZhWKHIuSxBrkH/QwO9NY7oJKeya707t0JbrcbPO9r6ukPgBQKKXwrJQMAg8vlhaxmPR2z2YqCgjIAQEpKHDiOg0TCw253BfXHojYRhBBC2gIKfFpA+virYE8dgi+LYuFGcO2OhGPoofO1kEhWh24QqlBI4XC4EROjBM/zNevr+BYK9K/q7HYLOF1iwqlTFfj22x3iGjsAMHhwNyx75wGkpiYgL68k6PrUJoIQQkhbQIFPCxhw043IcRuD9scr3EgzONBL74Syngah/syO2WyFXC6Dw+6C3hADqTRwYUOTyYpXlq4Wg526tTh79vyFR+eupDYRhBBC2jQKfFpA9voN6HDJZQAACbzooXOgb6wD7RQuSKV8TTuIYP7aHY9HQGlpJZKSjHj1lTUYdc1A9OuXCrlcBpfL3eA2D9QmghBCSFtHs7rqiNSsruU7V6FHOx7S8jwYtAokJhrE1ZClUr5W/63AthaMMRQWlqOqygaNRoVbJ/wb+/blNGu2Fc3WIoQQ0trQdPYmikTgA0DsxaXRqsTbTLGxMdDrNaiutmPd//Zg6r2jIZP5aoAYA7xeAcXFFSgpMSMlJQ5ZWfm0YjIhhBASQkM/v8/SvpKEk/82U1ZWPjQaFZKSjJBKJdiz5y9Mvec1zJz5Fq4ftwDFxSZUV9tRWFiGrKxcVFXZkJISRzU4hBBCSBhQxqeOSGV8/M51myk9vb9Yg0MrJhNCCCENQ7e6mijSgU9DUA0OIYQQ0ji0cvMFjFZMJoQQQiKDanwIIYQQ0mZQxqceWq022kMghBBCSAM19HObAp86/N+4wsLCKI+EEEIIIY2l1WqpuLmxkpOTQ37TtFotCgsLkZKSErXCZxKMfi7nJ/q5nJ/o53J+op9LeGi1Wpw6deqsx1DGJ4RzfdMsFgv9j3keop/L+Yl+Lucn+rmcn+jn0jwN+d5RcTMhhBBC2gwKfAghhBDSZlDg0whOpxPPPPMMnE5ntIdCaqGfy/mJfi7nJ/q5nJ/o59JyqLiZEEIIIW0GZXwIIYQQ0mZQ4EMIIYSQNoMCH0IIIYS0GRT4EEIIIaTNoMCnAebPn4/t27fDarXCZDKFPIYxFrRNmzathUfatjTk59KxY0d89913qK6uRmlpKV577TXIZLIWHinJzc0N+v1YtGhRtIfV5syYMQM5OTmw2+3YvXs3rrjiimgPqU3LyMgI+r0oKiqK9rBaPVq5uQHkcjm++uor7NixA/fee2+9x02ZMgXr168XH1dWVrbE8Nqsc/1ceJ7H2rVrUVpaiiuuuAJxcXH48MMPwXEcZs2aFYURt21PPfUUli9fLj6urq6O4mjanttvvx2vvvoqZs6cie3bt2PatGlYt24d+vTpg4KCgmgPr83Kzs7GNddcIz72er1RHE3bwWhr2DZ58mRmMplCPscYYzfddFPUx9gWt/p+LmPHjmUej4clJSWJ+yZOnMjsdjvTarVRH3db2nJzc9lDDz0U9XG05W3nzp3srbfeCth38OBB9u9//zvqY2urW0ZGBtu3b1/Ux9HWNrrVFUZvvPEGSktL8dtvv2HatGngOC7aQ2rTLr30UmRnZwekjjds2AClUokhQ4ZEcWRt07x581BWVoZ9+/Zh/vz5dMuxBclkMgwZMgQbN24M2L9x40ZcdtllURoVAYCLLroIhYWFyMnJwWeffYYuXbpEe0itHt3qCpMnn3wSmzdvht1ux6hRo/Dyyy8jPj4eCxcujPbQ2qzExESUlJQE7DObzXA6nUhMTIzSqNqm1157DXv37oXJZMIll1yCRYsWoUuXLrjvvvuiPbQ2IT4+HlKpNOj3oaSkhH4XomjXrl24++67cfToUbRv3x5PPvkkfv31V/Tt2xcVFRXRHl6rFvW0UzS2jIwMdi5DhgwJOOdst7rqbnPmzGFmsznq7/NC28L5c3nnnXfY+vXrg/Y7nU42ceLEqL/XC31rys/Kv02YMIExxpjRaIz6+2gLW1JSEmOMseHDhwfsnz9/Pjt06FDUx0ebb1Or1ayoqIjNnj076mNpzVubzfi88cYb+Pzzz896TF5eXpOvv3PnTsTGxiIhIQGnT59u8nXamnD+XIqLizFs2LCAfXq9HnK5POhfvqTxmvOz2rlzJwCge/fu+O2338I9NFJHWVkZPB5PUHYnISGBfhfOIzabDVlZWbjooouiPZRWrc0GPuXl5SgvL4/Y9QcNGgS73Q6z2Ryx12iNwvlz2bFjB5544gkkJiaiuLgYAHDttdfC4XBgz549YXmNtqw5P6tBgwYBAE3dbSFutxt79uzB6NGjsXr1anH/6NGjsWbNmugNjASQy+Xo3bs3fv7552gPpVVrs4FPY3Ts2BFGoxGdOnWCRCLBgAEDAAB//fUXrFYrbrjhBiQmJmLHjh2w2+1IT0/HwoUL8e6778LlckV59K3XuX4uGzduxMGDB/Hxxx9j7ty5MBqNeOmll7B8+XJYLJYoj77tGD58OIYPH47MzExUVlbi4osvxiuvvII1a9bQNOoWtHTpUnz88cfYvXs3duzYgfvvvx+dOnXCsmXLoj20NuvFF1/E999/jxMnTiAhIQFPPvkkdDodPvzww2gPrdWL+v22831buXJlyPqFESNGMABszJgxbO/evayqqopVV1ez/fv3s1mzZjGJRBL1sbfm7Vw/FwCsY8eO7Pvvv2dWq5WVlZWx//znP0wul0d97G1pGzRoENuxYwczmUzMZrOxQ4cOsYyMDKZSqaI+tra2zZgxg+Xm5jKHw8F2797NrrzyyqiPqS1vn332GSssLGROp5OdPHmSff3116x3795RH1dr37iaLwghhBBCWj1ax4cQQgghbQYFPoQQQghpMyjwIYQQQkibQYEPIYQQQtoMCnwIIYQQ0mZQ4EMIIYSQNoMCH0IIIYS0GRT4EEIIIaTNoMCHENImZWRkYN++fdEeBgBg5cqV+Pbbb6M9DELaBAp8CCHN0r59e7z66qs4duwY7HY7iouL8fPPP2PatGlQqVTRHl6TZGRkgDF21i01NbXR101NTQVjTOwrRwhpedSklBDSZF26dMH27dthNpsxf/58ZGVlQSqVokePHpg6dSpOnTqF77//PuS5UqkUHo+nhUfcMC+99FJA887ff/8d7777LpYvXy7uKy0tFb+WyWRwu90tOkZCSNNFvWEYbbTRdmFu69atYydOnGBqtfqcxzLG2LRp09jq1atZdXU1e+aZZxgANn36dPbXX38xp9PJDh8+zO68807xnNTUVMYYYwMGDBD3xcbGBjSjHTFiBGOMsauvvpr9/vvvzGq1su3bt7MePXoEvP68efNYcXExq6qqYu+99x5btGgR27dvX4PeZ25uLnvooYfExytXrmTffvste+yxx1hhYSHLzc0V3+NNN90UcK7JZGKTJ08Wn68tMzMz4HoPP/wwO3XqFCsrK2NvvPEGk0qlUf8Z00Zba9voVhchpEmMRiOuvfZavPnmm7DZbA06Z8GCBVizZg369euHFStW4Oabb8Zrr72Gl19+GWlpaXjnnXewcuVKjBw5stHjWbhwIR5++GEMHToUHo8HK1asEJ/729/+hgULFuCJJ57A0KFDUVRUhJkzZzb6NWobNWoUevfujdGjR+OGG25o0DkXX3yxeG5iYiImTJggPpeeno5u3bohPT0dkydPxpQpUzBlypRmjZEQElrUoy/aaKPtwtsuueQSxhhjN998c8D+0tJSZrFYmMViYS+88IK4nzHGli5dGnDsL7/8wt55552AfV988QX74YcfGND4jI//mOuuu44xxphCoWAA2Pbt29lbb70V8Do7duxoVsanqKiIyWSygOPOlfEJ9X7818vNzWU8zwd8Hz777LOo/5xpo621bZTxIYQ0C2Ms4PEll1yCgQMH4sCBA1AoFAHP7d69O+Bx7969sX379oB927dvR+/evRs9jv3794tfFxUVAQASEhLE19mxY0fA8XUfN1ZWVlZY63oOHDgAQRDEx0VFReL4CSHhQ4EPIaRJ/vrrLwiCgF69egXsz83NxfHjx2G324POsVqtQfvqBk4cx4n7/IEAx3Hi8zKZLOR4agch/vN5PnJ/4kK9F0EQAsYK1D/euuoGUYyxiI6fkLaKfqsIIU1SUVGBH3/8Ef/617+gVqubdI1Dhw7hiiuuCNh32WWX4dChQwDOzJxKSkoSnx84cGCTXmf48OEB++o+DofS0tKAsXbv3h0xMTHiY5fLBQCQSCRhf21CSMPQdHZCSJPNnDkT27dvx+7du/HMM89g//79EAQBF198MXr16oU9e/ac9fwXX3wRX375Jfbu3YvNmzdj/PjxmDBhAq655hoAgMPhwI4dO/DYY48hLy8P8fHxeP755xs9ztdeew0ffvghdu/ejV9++QX/+Mc/0LdvX+Tk5DTpfddny5Yt+Ne//oWdO3eC53ksXrxYDHYA4PTp07DZbBg7dixOnjwJh8OBqqqqsI6BEHJ2lPEhhDRZTk4OBg0ahE2bNmHRokX4888/sXv3bjz44IN46aWX8NRTT531/DVr1uChhx7C3LlzceDAAUybNg333HMPtm3bJh4zdepUyGQy7N69G6+99hqefPLJRo/zyy+/xLPPPovFixdjz549SE1Nxdtvv93o65zLww8/jIKCAvz000/49NNP8dJLLwXMePN6vZg1axamTZuGU6dOYc2aNWEfAyHk7Dj4qpwJIYQQQlo9yvgQQgghpM2gwIcQQgghbQYFPoQQQghpMyjwIYQQQkibQYEPIYQQQtoMCnwIIYQQ0mZQ4EMIIYSQNoMCH0IIIYS0GRT4EEIIIaTNoMCHEEIIIW0GBT6EEEIIaTMo8CGEEEJIm0GBDyGt0OTJk8EYq3cbMWJEtIcYUkZGRsA4rVYrCgoKsH79evzrX/+CRqMJOmflypXIzc1t1OskJSUhIyMDAwYMaNR5oV6LMYbXX3+9Udc5lxkzZmDy5MlB+1NTU8EYC/kcIaRhpNEeACEkcqZMmYLDhw8H7T948GAURtNwY8aMQWVlJeRyOZKTkzFq1CgsWbIEc+fOxfjx47F//37x2Oeeew6vvfZao66fnJyMZ555Bnl5efjzzz8bfF5TXqspZs6cibKyMnz44YcB+4uKijB8+HAcP3484mMgpLWiwIeQViw7Oxt79uxp9HlKpRIOh6PB+xtDpVLBbref9Zg9e/agvLxcfPzFF1/gjTfewLZt2/Ddd9+hR48ecLlcAICcnJxmjach/GNuidc6G5fLhV27dkV1DIRc6OhWFyFtnP9WzbRp03Dw4EE4nU5MnjxZvF02evRovP/++zh9+jTsdjsUCgU4jsPcuXNx6NAhOBwOlJSU4MMPP0RKSkrAtTMzM5GVlYUrr7wS27dvh9VqxYoVK5o0zv3792PhwoVITU3FxIkTxf2hbj/ddttt2LlzJ8xmM6xWK44fP473338fADBixAjs3r0bAPDBBx+It9UyMjLE61ksFqSlpWHDhg2oqqrC5s2b630tv/vvvx9HjhyBw+HAgQMHAsYInLmNV5f/+5yamgoAyM3NRVpaGkaOHCmOzf+a9d3quvzyy7Fp0yZUVVXBarVi+/btGDduXMjXGTlyJN566y2UlpairKwMq1atQlJS0lm+84S0LhT4ENKKSSSSoI3ng3/tb775ZsyYMQPPPvssxowZg59//ll8bsWKFXC73bjrrrtw6623wu124+2338aSJUvw448/4sYbb8RTTz2FsWPH4tdff0VcXFzAtZOSkvDJJ5/g008/xbhx4/DWW281+f189913AICrrrqq3mOGDx+OL774Ajk5OZg0aRKuv/56PPvss5BKfQnuvXv3YsqUKQB8t66GDx+O4cOH47333hOvIZfL8d1332HLli246aabxKCoPjfeeCNmzZqFp59+Grfddhvy8/Px+eef49Zbb230e7zllltw/Phx7N27VxzbLbfcUu/xV111FbZs2YLY2Fjce++9uOOOO2CxWPD999/j9ttvDzr+vffeg9vtxt///nc8+uijGDlyJD755JNGj5OQCxmjjTbaWtc2efJkVh+32x1wLGOMmUwmptfrQ17jgw8+CNjfs2dPxhhjb7zxRsD+iy++mDHG2PPPPy/uy8zMZIwxlp6e3qBxZ2RkMMYYi4uLC/m8QqFgjDG2du1acd/KlStZbm6u+HjOnDmMMcZ0Ol29rzNkyBDGGGOTJ08Oem7lypWMMcamTJkS8rnar+X//lmtVpaQkCDu43meHTx4kB09ejTovdX3s0pNTRX3ZWVlsczMzKBjU1NTg8b966+/suLiYhYTExPw+vv372cnTpwIep26P7dHHnmEMcZY+/bto/7/LW20tcRGGR9CWrG77roLQ4cODdiGDRsWdNyWLVtgNptDXmPVqlUBj9PT0wH4bhPV9vvvv+PgwYMYNWpUwP6KigpkZmY2/U3UwnHcOY/5/fffAQBffvkl/va3vyE5OblJr1X3fZ/N5s2bcfr0afGxIAj44osvcNFFFwXd/gsntVqNYcOG4euvv4bVag14/Y8//hgdO3ZEz549A87xZ838/IXi/ltthLR2FPgQ0oodOnQIe/bsCdj27t0bdFxRUVG916j7nP9WVqhzTp06FXSr62zXbiz/h/OpU6fqPebnn3/GTTfdBKlUio8++giFhYXIysrCpEmTGvw6VqsVFoulwccXFxfXu6/u9yOcDAYDeJ6v92cR6vVrF40DgNPpBOAr4CakLaDAhxASsui2vuf8H5yhCmKTk5NRVlbW4Gs31o033ggA2Lp161mP++6773DNNdcgNjYWI0aMwMmTJ/HZZ59h+PDhDXqdxo45MTGx3n3+75d/NpxcLg84Lj4+vlGvVZvJZILX6633ZwEg6OdBSFtHgQ8hpFG2bNkCALjzzjsD9g8dOhR9+vQRZ0CFW//+/TF//nzk5ubiyy+/bNA5LpcLP/30E+bNmwcAGDRoEIDwZzlGjRqFhIQE8THP85g4cSL++usvFBYWAgDy8vLE91Hb+PHjg67ndDobNDabzYZdu3ZhwoQJUCqV4n6O43DnnXeioKAAR48ebcpbIqTVonV8CGnF0tLSxNlMtR0/frzJmYCjR4/inXfewYMPPghBELBu3Tp07twZzz33HE6cOIFXXnmlucPGkCFDUFlZCZlMJi5geNddd+H06dMYP3483G53vecuWLAAHTp0wObNm3Hy5Eno9Xo89NBDcLlc2LZtGwDf+7fZbPjHP/6BQ4cOobq6GqdOnWrybbmysjJs2bIFzz33HKxWK2bOnInevXsHTGn/3//+h/Lycrz//vt4+umn4fF4MGXKFHTs2DHoev5bc7fffjtycnLgcDiQnZ0d8rUff/xx/Pjjj8jMzMRLL70El8uFmTNnIi0tDXfccUeT3g8hrV3UK6xpo4228G5nm9XFGGP33nuveCxjjL3++uv1XmPIkCFBz3Ecx+bOncsOHz7MnE4nO336NPvoo49YSkpKwHGZmZksKyurweP2z3zys9vtrLCwkK1fv549+OCDTKPRBJ1Td6bVuHHj2Nq1a1lBQQFzOBysuLiY/fDDD+zyyy8POG/ixIns4MGDzOl0MsYYy8jIEK9nsVhCjq++WV2vv/46mz59Ojt27BhzOp3s4MGD7I477gg6f+jQoeyXX35hFouFFRQUsIyMDDZ16tSgWV2dOnVi69evZ5WVlYwxJr5mqFldANjll1/ONm3axCwWC7NarezXX39l119/fYN+niNGjGCMMTZixIio/39LG20tsXE1XxBCCCGEtHpU40MIIYSQNoMCH0IIIYS0GRT4EEIIIaTNoMCHEEIIIW0GBT6EEEIIaTMo8CGEEEJIm0ELGIaQnJzcqD49hBBCCIk+rVZ71l5+AAU+QZKTk8Ul5gkhhBByYUlJSTlr8EOBTx3+TE9KSgplfQghhJALhFarRWFh4Tk/uynwqYfFYqHAhxBCCGllmlXcPGjQIKSlpYmPb7zxRnz77bdYuHAhZDJZswdHCCGEEBJOzQp83nnnHfTo0QMA0KVLF3z++eew2Wz429/+hiVLloRlgIQQEg0cx4Hn+bBtHMdF+y0RQtDMwKdHjx74448/AAB/+9vf8NNPP+Ef//gHpkyZgltvvTUc4yOEkBbHcRwKTp6E1+s952Z3Oht0XMHJkxT8EHIeaFaNj/9fRABwzTXX4IcffgAAFBQUID4+vvmjI4SQKOA4DinJyXjwg7chMFbvcRKex38mT8eDHy6DIAj1HsdzHF6fMgMcx4Gd5XqEkMhrVuCze/duPPnkk9i0aRNGjBiBGTNmAPDd9iopKQnLAAkhJFoExs4aqPiDonMeF/aREUKaqlm3uv7v//4PgwcPxhtvvIGFCxfi+PHjAIDbbrsNv/76a1gGSAghhBASLs3K+GRlZaF///5B++fOnQuv19ucSxNCSERwHHfOWhv/Lfxwa+h12TkySISQpovIOj5OpzMSlyWEkGbxFy2nJCe37OsC8AoC3G53g44vPHUKHTt0oOCHkAhoUuCTk5MT8AvZrVu3sA2IEEIipbFFy+A4IBzBB8dBwvPnLIIGqBCakEhrUj53ypQpuOeee8StuR577DH89ttvqKqqQklJCb799ltxfaDaMjIyUFhYCJvNhszMTPTp0yfgeblcjv/85z8oLS1FdXU11qxZg5SUlGaPjxDSuviLkevbzhYURfJ1I/nahBCfJgU+P/30U8DWXCNGjMCbb76J4cOHY/To0ZBKpdi4cSPUarV4zKOPPoo5c+bgX//6Fy6++GIUFxfjxx9/hEajEY959dVXccstt2DSpEm44ooroNFo8MMPP0Tsfj0hhBBCLizNrvHhOA7du3dHQkJCUIDx888/N+ga1113XcDje+65B6WlpRgyZIh4jf/7v//DwoUL8e233wIAJk+ejJKSEvz973/Hu+++C51Oh3vvvRd33XUXNm/eDAC48847UVBQgGuuuQYbN25s7lslhBBCyAWuWYHPsGHD8OmnnyI1NTVolgRjDFJp0y4fGxsLAKioqADgWxcoKSkpIHhxuVzYtm0bLrvsMrz77rsYMmQI5HJ5wDFFRUXIzs7GZZddVm/gI5fLoVAoxMdarbZJYyaERE9DZmoBkZutRQi5cDTrr8CyZcuwe/dupKWlwWg0wmAwiJvRaGzydZcuXYqff/4ZBw4cAAAkJiYCQNCiiCUlJeJziYmJcDqdMJvN9R4TyuOPP46qqipxKywsbPK4CSEtrzHtJRo6q4oQ0no1K+Nz0UUX4bbbbhMXLgyHN954A/3798cVV1wR9FzdGQ4NmfVwrmMWLVqEpUuXio+1Wi0FP4RcQBo6UwuIwGwtQsgFp1kZn127dqF79+7hGgv+85//4MYbb0R6enpA8FFcXAwAQZmbhIQEMQtUXFwMhUIBvV5f7zGhuFwuWCyWgI0QcuGhGVOEkIZoVuDz+uuv4+WXX8bkyZMxePBg9OvXL2Br7LUmTJiAq6++Gnl5eQHP5ebmoqioCKNHjxb3yWQyjBgxQmyNsWfPHrhcroBjEhMTkZaWRu0zCCGEEAKgmbe6Vq1aBQBYsWKFuI8xJt5eamhx85tvvom///3vuOmmm2CxWNC+fXsAQGVlJRwOBwDfVPX58+fj2LFjOHbsGObPnw+bzYZPP/0UAFBVVYX3338fL7/8MsrLy1FRUYGXXnoJWVlZ2LRpU3PeJiGEtLiGFGJTawtCGq9ZgU+XLl3CMoiZM2cCALZt2xawf8qUKfjwww8BAEuWLIFKpcJbb70Fg8GAXbt24dprr0V1dbV4/OzZs+HxePDll19CpVJh8+bNmDJlyjlXSiWEkPNFY9pbUGsLQhqvWYHPiRMnwjKIhkxDBYAFCxZgwYIF9T7vdDoxa9YszJo1KyzjIoSQFtfA9hbU2oKQpml04DN+/HisW7cOHo8H48ePP+ux33//fZMHRgghbZlwjttYlMcmpGkaHfisXr0aiYmJKC0txerVq+s9rjkLGBJCCGkYqgUipHEaHZlIJJKQXxNCCGk5VAtESNM0KSVz7733YsuWLcjNzQ33eAghRNSQVhRttg0F1QIR0iRNCnxee+01KJVKFBYWIjMzE5mZmdiyZQv+v707j4qy3v8A/h5kERDcEAUTksUkS+1CqbiAEmmraeV2UkTP1fCmpbkk1y4ux7RIBTW10ktl12vWQfR2rksamKJp2HIF7SqKyC6yCLKjn98f/pjLCMPMMAMDM+/XOZ9T88yzfOaZcfjM9/l+n29GRoah8yMiM1U3FUVvV1djp9KmsS8QkW6aVfh06dIFQ4cORUBAAEaPHo2PP/4YHTt2RHp6On744QdlMZSTk2PofInIBGjbkqPNVBSchoKIdNGswqe2thanTp3CqVOnsHbtWlhaWmLo0KEYPXo0AgMDMXXqVNjY2MDKysrQ+RJRO6drS47GFg0WO0SkA4MMu+rQoQOsra1hY2MDGxsbWFpasv8PETVK20lF2ZJDRC2hWYWPjY0N/P39ERgYiDFjxsDPzw/Xrl3Djz/+iK1bt+LEiRO8zEVETWJLDhEZQ7MKn+LiYuTl5eHgwYOIjo7GiRMnkJ+fb+jciIiIiAyqWYXP77//jsGDByMgIAAignv37iEhIQGFhYWGzo+IiIjIYJp1A4yhQ4eiW7duWLp0KSoqKrB06VLk5OTgwoUL2LJlC1599VX06NHD0LkSERER6aXZnZvLy8tx5MgRHDlyBADQqVMnjBw5EsHBwfjss8/QqVMnjuoiIiKiNkXvUV0KhQJPPvkkAgMDMXr0aAwfPhz29vZIT083RH5EREREBtOswsfPz095z54RI0agU6dOyMzMREJCAhYsWID4+HgWPkRERNTmNKvwOXv2LHJycpCQkIBFixYhISEBV69eNXRuRERERAbVrMLHx8cHly9fNnQuRERERC2qWaO6WPQQERFRe2SQKSuIiKhts7DQ/DtXNNxNm8gUNKvFh4iI2gcFgLv37qGmpgZ3795tMjIyM6FQKIydMlGL0rnFx8HBAaWlpS2RCxERGZpCgQ4WFpj/xQ7cu3dP7WoWCgW2zAyDQqFgqw+ZNJ1bfIqKipR3ZT5+/Dg6d+5s8KSIiMiw6iaFVRecFJbMhc6Fz507d9C9e3cAQGBgIO/OTERERO2Gzpe6jh07hvj4eFy6dAkAsH//flRXVze6blBQkH7ZERERERmQzoXP66+/jpCQEHh6eiIgIAApKSkoLy9vidyIiKiVaTP6C+AIMGq/dC58Kisr8cknnwC4P3XFsmXLcPv2bYMnRkREraf+6C9tZGVnw93NTWPxwwKJ2hq97uMzZswYQ+VBRETGpOXoLwDooFBg04y5qK2t1bjbrOxs9HnoIRY/1GbofQPD6dOnY8mSJfD29gZw/67OkZGR+Oqrr/ROjoiIWtc9LVpo7nGIPLVjehU+CxcuxJo1a7B161YkJiZCoVBg+PDh2LFjB5ycnBAVFWWgNImIqK3RVCQ13W5EZBx6FT7z589HWFgYdu/erVx28OBBpKSkYOXKlSx8iMyMQqHQeOdfbTvPEhG1BL2+gVxcXHD69OkGy0+fPg0XFxed9jVy5EgcPHgQWVlZEBGMHz++wToRERHIyspCeXk54uPj8eijj6o8b21tjc2bNyM/Px937tzBgQMH0Lt3b91eFBE1i0KhQEZmpsZpEbTtPEtE1BL0avFJTU3FpEmTsG7dOpXlkydPxpUrV3Tal729PX7//XfExMQgNja2wfNLly7FokWLMHPmTFy+fBkrVqzA999/j0ceeQR37twBAERFReHFF1/ElClTUFBQgA0bNuC7776Dr6+vxs56RKQfhUKB3q6umP/59ibvAtzBwgKbQ94AFAqA/T6IqJXpVfhERETg66+/xqhRo5CYmAgRwYgRIxAUFIRJkybptK/Dhw/j8OHDap9/++23sXbtWuzfvx8AEBISgry8PEybNg2ffvopHB0dMXv2bEyfPh3Hjx8HcP+eQxkZGXj66adx9OjR5r9QIjOmzeUr4H+XsDT2+2CxQ0RGpNelrtjYWAwZMgS3bt3Cyy+/jIkTJ+LWrVt46qmnEBcXZ6AUgb59+8LFxUWleKmursaJEyfg7+8PAPD19YW1tbXKOjk5OUhOTlauQ0S60fbyFS9hEVF7ofdw9l9++QXTp083RC5q9erVCwCQl5ensjwvLw/u7u7KdaqqqlBcXNxgnbrtG2NtbQ0bGxvlYwcHBwNlTdT+aXv5CuAlLCJqH9rV8IoHm8+1uTeEpnWWL1+OkpISZWRlZRkkVyJTomlmb87uTUTtRbsofHJzcwGgQcuNs7OzshUoNzcXNjY26NKli9p1GrNu3To4Ojoqg6PAiIiITFe7KHzS0tKQk5OD4OBg5TIrKysEBAQoh9OfP38e1dXVKuv06tULjz32WKND7utUV1ejtLRUJYiIiMg06d3Hx1Ds7e3h5eWlfNy3b18MGjQIhYWFyMjIQFRUFMLDw3HlyhVcuXIF4eHhKC8vx549ewAAJSUl2LVrFzZs2ICCggIUFhbio48+woULF3Ds2DFjvSwiIrOnzU0rOZkptZY2U/j4+fkhISFB+XjTpk0AgM8//xyhoaH48MMPYWtri23btqFr1644e/YsnnnmGeU9fID7U2jU1tZi3759sLW1xfHjxzFz5kzew4eIyAh0mfGdk5lSa2mRwicsLAxOTk5Ys2aN1tucOHFC471CVq1ahVWrVql9vqqqCgsWLMCCBQu0Pi6RqdH2vjv8hU0tjpOZUhvUIn18XnnlFcycObMldk1ETdDlvjsZmZlaFUhE+tI0KpAjAqk1tUiLz9NPP90SuyUiDbS97w5/YRORuWozfXyIyHA0ThvRirkQEbUlehc+rq6uGD58OJydnRv03N+yZYu+uyciIiIyGL0Kn5kzZ2LHjh2orq5GQUGByi9MEWHhQ0REWuOwd2oNehU+q1evxurVq7Fu3Tp+EImIqFk47J1ak16Fj52dHfbu3csPIFE7pekXtja/wIn0xmHv1Ir0+lbbtWsXXnvtNUPlQkStpP4v7KaGvGvzC5zIUDjsnVqDXi0+y5cvx3fffYdx48bhwoULDb4k33nnHb2SI6IWouUv7A4WFtgc8gagUAD8o0NEJkCvwic8PBxjx47Ff//7XwBo0LmZiNo2jcPe+e+YiEyMXoXPokWLMGvWLHzxxReGyoeIiIioxejVx6eqqgqJiYmGyoWIiIioRelV+ERHR2P+/PmGyoWImqBQKGBhYaExiIhIPb0udT311FMYM2YMXnjhBaSkpDTo3PzKK6/olRwR3Vc3+WhvV1djp0JE1K7pVfgUFxcjNjbWULkQkRraTj7KUVhERE3Tq/CZNWuWofIgIi1wFBaR9jfW5PQW1BjOzk5ERO2CLlNbAPent3B3c9NY/LBAMi96FT7Xrl1r8sPi6empz+6JTJ5CoYBCodC4HjstE0HrG28CQAeFAptmzEVtba3G3XL+L/OiV+ETFRWl8tjKygpPPPEExo0bh8jISH12TWTy2GGZqHk0XfIFgHuc/4vU0Kvw2bx5c6PL582bBz8/P312TWTytO2wDLDTMlFzaewX14q5UNvQIu3nhw4d4lB2Ii1pmpiRkzMSERlOi3RufvXVV1FYWNgSuyZqF7Tpu8N+O0RErU+vwueXX35RaUJUKBTo1asXevTogXnz5umdHFF7xL47RERtl16FT1xcnMrje/fuIT8/HwkJCcoZ24nMDW82SETUdulV+KxevdpQeRCZHN5skKj90ObSM+/3Yxp4A0MiHbDvDpFp0eWmiLzfj2loVuFz9+5dre6EaWVl1aykiFqbNgWNQqFA+o0b7LtDZEp4vx+z06zCZ8KECWqf8/f3x/z587W6Gy1RW6BrZ2T23SEyPdre74eXxNq/ZhU+Bw8ebLDskUcewbp16/Diiy/iH//4B9577z29kyNqDbp2Rr4HsO8OkZnhJTHToXcfHxcXF6xatQohISE4cuQIBg8ejJSUFEPkRqQXXefBYmdkIlKLl8RMRrN7YTo6OmL9+vVITU3FgAEDEBQUhJdeeqlNFD1hYWG4du0aKioqkJSUhBEjRhg7JdKCQqGAhYWFxtCmmKm7fHX37l2Noe1Mz0REmu60XvcDSZvvMm2/z8iwmtXis2TJEixbtgy5ubmYOnVqo5e+jGXSpEmIiorCvHnzkJiYiLlz5+LQoUN49NFHkZGRYez0SA1d+tlkZWfD3c2tyV9TFhYWnAeLiFqdLpfEAO2+z3TB/kWaNavwWb9+PSoqKpCamoqQkBCEhIQ0up4x5utatGgRdu3ahV27dgEAFi5ciLFjxyIsLAzh4eGtno+p0vYyEqDdP0St+9koFNg0Yy5qa2u1OrZWszjzS4KIDEXLS2KAbt9n1bW1sLbU/Cfb0IUUYHrFVLMKny+//LJNngQrKyv4+vpi/fr1KsuPHj0Kf39/I2X1P7oUC22ZrsO6denop7GfjZZfKmzFISJj0upHl47fZxrXa4FCCjC9VqlmFT6hoaGGzsMgnJycYGlpiby8PJXleXl56NWrV6PbWFtbw8bGRvnYwcFB5b+GolAocOmPP+Dq4qJxXV0+kNqua+j1AGDpnl1ateR8OG02unTp0nSHwP/vYGxnba1xZBUAdLSy0vhFodyfpl9dWq5rrPWYI3NsS+sxR8PnqO33mTbrdbCwwNJ//h2i4fv2g6mzNK4H3O+s/f6UUIMWU9k5OfDp39/gxY+2f7cVAEzm57CLiwuys7MxbNgw/PTTT8rl4eHhmD59Onx8fBpsExERgZUrV7ZilkRERNRSevfujezsbLXPm9SUFbdu3UJtbW2D1h1nZ+cGrUB11q1bh40bN6os69atGwoLC1ssz7bAwcEBWVlZ6N27N0pLS42djtnh+Tcunn/j4vk3LlM+/w4ODk0WPYCJFT41NTU4f/48goODVWaODw4OxoEDBxrdprq6GtXV1SrLTO2D0JTS0lKzer1tDc+/cfH8GxfPv3GZ4vnX5vWYVOEDABs3bsTu3buRlJSEM2fOYM6cOXBzc8OOHTuMnRoREREZmckVPvv27UP37t3xt7/9DS4uLkhOTsZzzz2HGzduGDs1IiIiagOEYX5hbW0tERERYm1tbfRczDF4/nn+zTl4/nn+jRkmNaqLiIiIqCnNnquLiIiIqL1h4UNERERmg4UPERERmQ0WPkRERGQ2WPgQ3N3dsXPnTly7dg3l5eVITU3FypUrYWVlZezUzEZ4eDgSExNRVlaGoqIiY6dj8sLCwnDt2jVUVFQgKSkJI0aMMHZKZmHkyJE4ePAgsrKyICIYP368sVMyK++++y7OnTuHkpIS5OXlYf/+/ejXr5+x02p1LHwI/fv3h4WFBebOnYsBAwZg4cKFeOONN/D+++8bOzWzYW1tjW+++Qbbt283diomb9KkSYiKisLatWvxxBNP4OTJkzh06BD69Olj7NRMnr29PX7//Xe8+eabxk7FLAUEBODjjz/G0KFDERwcDEtLSxw9ehR2dnbGTq3VGX1MPaPtxeLFi+Xq1atGz8PcIiQkRIqKioyehynHTz/9JNu2bVNZdvHiRXn//feNnps5hYjI+PHjjZ6HOYeTk5OIiIwcOdLoubRmsMWHGtW5c2eTn6iVzI+VlRV8fX1x9OhRleVHjx6Fv7+/kbIiMo7OnTsDgNl917PwoQY8PDwwf/58zm9GJsfJyQmWlpbIy8tTWZ6Xl4devXoZKSsi49i4cSNOnjyJlJQUY6fSqlj4mLCIiAiISJPh6+urso2LiwsOHz6Mb775Brt27TJS5qahOeefWoeIqDxWKBQNlhGZsq1bt2LgwIGYOnWqsVNpdSY3SSn9z9atW7F3794m17l+/bry/11cXBAfH6+c1Z70o+v5p5Z369Yt1NbWNmjdcXZ2btAKRGSqNm/ejJdeegmjRo1CVlaWsdNpdSx8TFhBQQEKCgq0WtfV1RXx8fE4f/48QkND+evXAHQ5/9Q6ampqcP78eQQHByMuLk65PDg4GAcOHDBeYkStZMuWLZgwYQICAwPN9ocXCx+Ci4sLEhIScOPGDSxevBg9evRQPsdfwa2jT58+6NatG9zc3NChQwcMGjQIAJCamoqysjIjZ2daNm7ciN27dyMpKUnZuunm5sY+ba3A3t4eXl5eysd9+/bFoEGDUFhYiIyMDCNmZh4+/vhjTJs2DePHj0dpaSl69uwJALh9+zYqKyuNnF3rMvrQMoZxIyQkRNQxdm7mEjExMY2e/4CAAKPnZooRFhYmaWlpUllZKUlJSWY3nNdYERAQ0OjnPCYmxui5mUOoExISYvTcWjMU//8/RERERCaPo7qIiIjIbLDwISIiIrPBwoeIiIjMBgsfIiIiMhssfIiIiMhssPAhIiIis8HCh4iIiMwGCx8iIiIyGyx8iNqAiIgI/Prrr8rHMTEx2L9/v177NMQ+2rtu3bohLy8P7u7uLXqctLQ0vPXWWy16DHVCQkJQVFRklGMbS48ePXDz5k24uroaOxVqp4x++2gGoy1G/Wkkqqur5erVqxIZGSl2dnYGP1ZERIT8+uuvyseOjo7SuXNnrbZ1d3cXEZFBgwapLNdlH/qEOpMnTzb6exgZGSk7d+5scK4aM2TIEI37CwkJkaKiogbLnZycxNbWtsVfT1pamrz11lsqyzp27Cg9evQw+rmuCwcHB1m9erUkJydLeXm53Lp1S86dOydLliyRLl26NLpNfHy82vdFRCQtLa3BNhs2bJDPPvvM6K+X0f6Ck5QSNeHQoUMIDQ2FlZUVRo4ciZ07d8Le3h7z5s1rsK6lpSVqa2sNctySkpI2sQ9tzZw5E4cPH1ZZVlxc3Oi6FhYWEBGIiMpyKysr1NTU6Hxsddt17NgRs2fPxnPPPdfguaCgIKSkpKgsKygo0PnYdW7dutXsbfVVWVnZZiaY7Nq1K06dOgVHR0e89957OH/+PKqrq+Hl5YVp06Zh2rRp2LZtW4PtJk6cCGtrawD3J+z9+eefVd6ju3fvNtgmJiYG586dw5IlS9R+1ojUMXr1xWC0xYiJiZH9+/erLPv0008lOztbgP+10oSGhsrVq1fl7t27Atxvafnkk08kLy9Pbt++LcePH5eBAweq7GfZsmWSm5srJSUlsnPnTlm3bp1Ki8+Dx1YoFLJ06VK5cuWKVFZWSnp6uoSHhwvQsMUlPj6+0X1YW1tLdHS05OXlSUVFhZw8eVL8/PyUz9dNIDlmzBj5+eefpaysTBITE6Vfv35NnicRkfHjx6t9vq6V5Pnnn5eUlBSpqamRhx9+WNLS0uSvf/2rxMTESHFxsXz++ecCQCZOnCjJyclSWVkpaWlpsmjRIpX9qdvuwZgwYYLcvHlTZZm61rH6MXDgQPnhhx+kpKREbt++LUlJSeLr69voBJsRERHKnOq3xIiIzJkzR/71r39JWVmZXLx4UYYOHSqenp4SHx8vd+7ckdOnT4uHh4dyGw8PD4mLi5Pc3FwpLS2Vc+fOSVBQkPL5xlpF6p/f+q/hjTfekNTUVKmqqpI//vhDXn/99Qbv2ezZsyU2NlbKysrk8uXL8uKLL+r9b2b79u1SWloqrq6uzd6HNu9RXVy7dk1CQ0Nb9XuBYRJh9AQYjDYZjRU+0dHRkp+fL8D9wqe0tFQOHTokgwcPlscff1wAyMmTJ+XAgQPi6+srXl5eEhkZKfn5+dK1a1cBIK+99ppUVlbK7NmzpV+/frJmzRq5fft2k4XP+vXrpaCgQGbMmCEeHh4yfPhwmT17tgAQPz8/ZcHSs2dP5XEe3EdUVJRkZmbKuHHjxMfHR2JiYqSgoEC5ft0f9jNnzsioUaPEx8dHTpw4IadOnWryPGlT+FRVVcmpU6dk2LBh0q9fP7Gzs5O0tDQpLi6Wd955Rzw9PcXT01P+9Kc/SW1traxYsUK8vb0lJCREysrKVGaPbmy7xo67adMm+fe//62yTJs/qhcuXJAvv/xSHnnkEfHy8pJXX31VBg4cKFZWVrJgwQIpLi6Wnj17Ss+ePcXe3l6Z04OFT0ZGhrz22mvi7e0tsbGxcu3aNTl27Jg888wz0r9/fzl9+rRKfgMHDpQ5c+bIY489Jl5eXrJmzRopLy+XPn36CADp2rWr3LhxQ1asWKE8ft35rV/4vPzyy1JVVSVhYWHi7e0tCxculJqaGgkMDFTJ78aNGzJlyhTx9PSUqKgoKSkpUX4WmhMKhUIKCwtl+/btev2706Xw2bt3r/z97383+ncFo92F0RNgMNpkPFg4PPnkk5Kfny979+4V4H7hU1VVJU5OTsp1Ro8eLcXFxWJtba2yrytXrsif//xnASCJiYmybds2lefPnDmjtvDp1KmTVFRUKAudB0PdH4r6+7Czs5OqqiqZOnWq8nlLS0vJzMyUxYsXC6Da4lO3zrPPPisiIjY2NmrPk4hIeXm5lJaWqkTfvn0FuP+HWUQatHqlpaVJbGysyrKvvvpKjhw5orLsgw8+kOTk5Ca3ayz279+v0r+n/rkqKytrkK+FhYUAkNu3b8uMGTMa3ae6Pj6NFT6rV69WPh4yZIiIiErrxOTJk6W8vLzJ15CcnCx/+ctf1B6nsZxOnToln3zyico6X3/9tXz33Xdq87Ozs5O7d+/K2LFjm/3vxdnZWURE3n77bZXlSUlJynO8Z88ejfvRpfDZsGGD/PDDD83OmWGewVFdRE144YUXUFpaioqKCpw5cwY//vgj5s+fr3w+PT1dpX+Hr68vOnXqhIKCApSWliqjb9++8PT0BAD4+PjgzJkzKsd58HF9Pj4+6NixI44fP97s1+Hp6Qlra2skJiYql9XW1uLcuXPw8fFRWfc///mP8v9zcnIAAM7Ozk3uf+HChRg8eLBKZGRkKJ+vqqpS2W+dpKQklcc+Pj4qOQJAYmIivL29YWFhoXa7xtja2qrt+zJ58uQG+d67dw8AsHHjRuzcuRPff/89li1bBg8PD43Hakz915uXlwcAuHDhgsoyW1tbODg4AADs7OzwwQcfICUlBUVFRSgtLUX//v3h5uam03HVncOm3ufy8nKUlpaqfZ+XL1+u8nnu06eP2uPLA323JkyYgMGDB+PIkSOwtbXV6bVoUlFRATs7O4Puk0wfOzcTNSE+Ph5hYWGoqalBdnZ2g87LZWVlKo8tLCyQk5ODwMDABvtqbgfMioqKZm1Xn0KhANDwj5JCoWiwrH5H4brn6hcdjcnNzcXVq1fVPq/uNTx4/hrLpy73prZrzK1bt9C1a9dGn8vIyFCb76pVq7Bnzx48//zzePbZZ7Fq1SpMmTIFcXFxGo9ZX2PnsalzGxkZibFjx2Lx4sVITU1FRUUFvv32W2WnX13o+j7XbaPufd6xYwf27dunfJydnd1gnfz8fBQVFaF///4qy+sK4NLSUnTp0kXr16CNbt26IT8/36D7JNPHFh+iJpSVleHq1au4ceOGViO2fvnlF/Tq1Qu1tbW4evWqStSNGrp06RKGDh2qst2Dj+u7cuUKysvLERQU1Ojz1dXVAIAOHTqo3UdqaiqqqqowYsQI5TJLS0v4+fnh0qVLGl9Xa7l48aJKjgDg7++Py5cvK1tktPXrr7/i0UcfbVYeV65cQVRUFMaOHYvY2FiEhoYCuH+umzrP+hg5ciQ+//xzxMXFITk5Gbm5uXj44YdV1tHm+JcuXWr0HOrzPhcVFal8lhsbZSUi2LdvH15//fVWu7/OY489pnL/KyJtsPAhMqBjx47hzJkziIuLwzPPPAN3d3cMGzYMa9asga+vLwAgOjoas2bNQmhoKLy9vbFy5UoMGDBA7T6rqqrwwQcf4MMPP8T06dPh4eGBIUOGYNasWQCAmzdvory8HOPGjYOzszMcHR0b7KO8vBzbt29Xtir4+Pjgs88+g52dHXbt2qX36+7SpQt69uypEs25BLFhwwYEBQVhxYoV8Pb2xowZM/Dmm2/io48+0nlfR44cwYABAxptZejevXuDfG1sbNCxY0ds2bIFAQEBcHNzg7+/P5588kll0XD9+nU4ODhgzJgx6N69u0Ev3aSmpmLixIkYNGgQBg4ciD179jRogbl+/TpGjRoFV1dXdO/evdH9REZGYubMmZg7dy68vLywcOFCTJw4sVnnUFfh4eHIysrC2bNnERoaiscffxweHh54+eWXMWzYMJWC6f3338cXX3yh1X7r3oP6BZWtrS18fX1x9OhRg78OMn1G72jEYLTFaGxUV/148KaDddGpUyeJjo6WzMxMqaqqkvT0dNm9e7c89NBDynWWL18uN2/elJKSEomJiZH169drHM4eHh4uaWlpUlVVJdevX5d3331X+fzs2bMlPT1damtr1Q5nt7GxkejoaLl582aTw9nr3/Rw0KBBIiLi7u6u9jyos2zZMgG07xBcF3XD2ete5zvvvKPVdo3F6dOnZc6cOcrHTd3AcPLkyWJlZSV79uyR9PR0qayslMzMTNm8ebNK5+5t27ZJfn6+iDQ9nL3+SLfGOuw+eL7d3d3l+PHjUlZWJunp6TJv3jyJj4+XTZs2KbcZMmSI/Pbbb1JRUSEi+g1nf3AkXlFRkcroueaGo6OjrF27Vi5evCgVFRVSVlYmv/32m6xatUpl1FhMTIzys1o/mjpX9T+HU6ZMkUuXLrXa9wHDpMLoCTAYDEaLxLPPPispKSmiUCiMngvDsHH27FmVUYoMhrbBzs1EZLIOHToEb29v9O7dG5mZmcZOhwykR48e+Pbbb/HPf/7T2KlQO6TA/QqIiIiIyOSxczMRERGZDRY+REREZDZY+BAREZHZYOFDREREZoOFDxEREZkNFj5ERERkNlj4EBERkdlg4UNERERmg4UPERERmY3/A8Oms19+SasQAAAAAElFTkSuQmCC",
      "text/plain": [
       "<Figure size 640x480 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fn.evaluate_model(y_train, y_val, y_test, y_train_pred, y_val_pred, y_test_pred)"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "1a17012b",
   "metadata": {},
   "source": [
    "Previous example: [/examples/time_series/cases/6.ipynb](https://github.com/serhatsoyer/py4ML/blob/main/examples/time_series/cases/6.ipynb)  \n",
    "Next example: [/examples/time_series/cases/8.ipynb](https://github.com/serhatsoyer/py4ML/blob/main/examples/time_series/cases/8.ipynb)"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3.10.6 ('study')",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.10.8 (main, Nov 24 2022, 08:08:27) [Clang 14.0.6 ]"
  },
  "vscode": {
   "interpreter": {
    "hash": "29d06b973d1ddb34db3279b24f9b5152402e688db937648b736e855ec4de60c0"
   }
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}