Github_Example.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "name": "Github_Example.ipynb",
      "provenance": [],
      "include_colab_link": true
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    },
    "accelerator": "GPU"
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "view-in-github",
        "colab_type": "text"
      },
      "source": [
        "<a href=\"https://colab.research.google.com/github/muriloasouza/Deep-Learning/blob/master/Github_Example.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "ghH5Q58FoRpc",
        "colab_type": "code",
        "outputId": "3557719c-81b7-4d85-b01f-e98a8963cfc3",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 68
        }
      },
      "source": [
        "# Install tensorflow 2.x in Colab\n",
        "\n",
        "try:\n",
        "  # %tensorflow_version only exists in Colab.\n",
        "  %tensorflow_version 2.x\n",
        "except Exception:\n",
        "    pass\n",
        "  \n",
        "# Import libraries  \n",
        "from __future__ import absolute_import, division, print_function, unicode_literals\n",
        "\n",
        "#Power data classification/regression with CNN\n",
        "import numpy as np\n",
        "import tensorflow as tf\n",
        "from tensorflow import keras\n",
        "import pandas as pd\n",
        "import time as time\n",
        "import matplotlib.pyplot as plt\n",
        "import pydot\n",
        "import csv as csv\n",
        "import keras.backend as K\n",
        "import itertools\n",
        "%matplotlib inline\n",
        "print(\"TensorFlow version:\",tf.__version__)"
      ],
      "execution_count": 1,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "TensorFlow 2.x selected.\n",
            "TensorFlow version: 2.1.0-rc1\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "Using TensorFlow backend.\n"
          ],
          "name": "stderr"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "6R3NP2r6o6y1",
        "colab_type": "code",
        "outputId": "2ffe7fc8-d41f-400e-c1a0-2eb3c3e4d5c2",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1000
        }
      },
      "source": [
        "# Downloading data files \n",
        "!wget https://raw.githubusercontent.com/muriloasouza/Deep-Learning/master/Input.txt\n",
        "!wget https://raw.githubusercontent.com/muriloasouza/Deep-Learning/master/Labels.txt\n",
        "!wget https://raw.githubusercontent.com/muriloasouza/Deep-Learning/master/Input_Train.txt\n",
        "!wget https://raw.githubusercontent.com/muriloasouza/Deep-Learning/master/Labels_Train.txt\n",
        "!wget https://raw.githubusercontent.com/muriloasouza/Deep-Learning/master/Input_Valid.txt\n",
        "!wget https://raw.githubusercontent.com/muriloasouza/Deep-Learning/master/Labels_Valid.txt\n",
        "  \n",
        "def data_set_size(csv_file):\n",
        "    with open(csv_file) as csvfile:\n",
        "        csv_rows = 0\n",
        "        for _ in csvfile:\n",
        "            csv_rows += 1\n",
        "    return csv_rows-1            #Remove header from count and return\n",
        "\n",
        "labels = \"./Labels.txt\"\n",
        "labels_train = \"./Labels_Train.txt\"\n",
        "labels_valid = \"./Labels_Valid.txt\"\n",
        "total = data_set_size(labels)\n",
        "total_train = data_set_size(labels_train)\n",
        "total_valid = data_set_size(labels_valid)\n",
        "print(\"Total Samples:\",total)\n",
        "print(\"Total Training Samples:\",total_train)\n",
        "print(\"Total Validation Samples:\",total_valid)"
      ],
      "execution_count": 2,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "--2020-01-09 11:37:31--  https://raw.githubusercontent.com/muriloasouza/Deep-Learning/master/Input.txt\n",
            "Resolving raw.githubusercontent.com (raw.githubusercontent.com)... 151.101.0.133, 151.101.64.133, 151.101.128.133, ...\n",
            "Connecting to raw.githubusercontent.com (raw.githubusercontent.com)|151.101.0.133|:443... connected.\n",
            "HTTP request sent, awaiting response... 200 OK\n",
            "Length: 892480 (872K) [text/plain]\n",
            "Saving to: ‘Input.txt.4’\n",
            "\n",
            "\rInput.txt.4           0%[                    ]       0  --.-KB/s               \rInput.txt.4         100%[===================>] 871.56K  --.-KB/s    in 0.02s   \n",
            "\n",
            "2020-01-09 11:37:31 (43.2 MB/s) - ‘Input.txt.4’ saved [892480/892480]\n",
            "\n",
            "--2020-01-09 11:37:35--  https://raw.githubusercontent.com/muriloasouza/Deep-Learning/master/Labels.txt\n",
            "Resolving raw.githubusercontent.com (raw.githubusercontent.com)... 151.101.0.133, 151.101.64.133, 151.101.128.133, ...\n",
            "Connecting to raw.githubusercontent.com (raw.githubusercontent.com)|151.101.0.133|:443... connected.\n",
            "HTTP request sent, awaiting response... 200 OK\n",
            "Length: 86669 (85K) [text/plain]\n",
            "Saving to: ‘Labels.txt.4’\n",
            "\n",
            "Labels.txt.4        100%[===================>]  84.64K  --.-KB/s    in 0.003s  \n",
            "\n",
            "2020-01-09 11:37:35 (30.7 MB/s) - ‘Labels.txt.4’ saved [86669/86669]\n",
            "\n",
            "--2020-01-09 11:37:38--  https://raw.githubusercontent.com/muriloasouza/Deep-Learning/master/Input_Train.txt\n",
            "Resolving raw.githubusercontent.com (raw.githubusercontent.com)... 151.101.0.133, 151.101.64.133, 151.101.128.133, ...\n",
            "Connecting to raw.githubusercontent.com (raw.githubusercontent.com)|151.101.0.133|:443... connected.\n",
            "HTTP request sent, awaiting response... 200 OK\n",
            "Length: 581395 (568K) [text/plain]\n",
            "Saving to: ‘Input_Train.txt.4’\n",
            "\n",
            "Input_Train.txt.4   100%[===================>] 567.77K  --.-KB/s    in 0.02s   \n",
            "\n",
            "2020-01-09 11:37:38 (36.3 MB/s) - ‘Input_Train.txt.4’ saved [581395/581395]\n",
            "\n",
            "--2020-01-09 11:37:41--  https://raw.githubusercontent.com/muriloasouza/Deep-Learning/master/Labels_Train.txt\n",
            "Resolving raw.githubusercontent.com (raw.githubusercontent.com)... 151.101.0.133, 151.101.64.133, 151.101.128.133, ...\n",
            "Connecting to raw.githubusercontent.com (raw.githubusercontent.com)|151.101.0.133|:443... connected.\n",
            "HTTP request sent, awaiting response... 200 OK\n",
            "Length: 54987 (54K) [text/plain]\n",
            "Saving to: ‘Labels_Train.txt.4’\n",
            "\n",
            "Labels_Train.txt.4  100%[===================>]  53.70K  --.-KB/s    in 0.002s  \n",
            "\n",
            "2020-01-09 11:37:41 (30.0 MB/s) - ‘Labels_Train.txt.4’ saved [54987/54987]\n",
            "\n",
            "--2020-01-09 11:37:44--  https://raw.githubusercontent.com/muriloasouza/Deep-Learning/master/Input_Valid.txt\n",
            "Resolving raw.githubusercontent.com (raw.githubusercontent.com)... 151.101.0.133, 151.101.64.133, 151.101.128.133, ...\n",
            "Connecting to raw.githubusercontent.com (raw.githubusercontent.com)|151.101.0.133|:443... connected.\n",
            "HTTP request sent, awaiting response... 200 OK\n",
            "Length: 145934 (143K) [text/plain]\n",
            "Saving to: ‘Input_Valid.txt.4’\n",
            "\n",
            "Input_Valid.txt.4   100%[===================>] 142.51K  --.-KB/s    in 0.005s  \n",
            "\n",
            "2020-01-09 11:37:44 (27.5 MB/s) - ‘Input_Valid.txt.4’ saved [145934/145934]\n",
            "\n",
            "--2020-01-09 11:37:48--  https://raw.githubusercontent.com/muriloasouza/Deep-Learning/master/Labels_Valid.txt\n",
            "Resolving raw.githubusercontent.com (raw.githubusercontent.com)... 151.101.0.133, 151.101.64.133, 151.101.128.133, ...\n",
            "Connecting to raw.githubusercontent.com (raw.githubusercontent.com)|151.101.0.133|:443... connected.\n",
            "HTTP request sent, awaiting response... 200 OK\n",
            "Length: 14018 (14K) [text/plain]\n",
            "Saving to: ‘Labels_Valid.txt.4’\n",
            "\n",
            "Labels_Valid.txt.4  100%[===================>]  13.69K  --.-KB/s    in 0s      \n",
            "\n",
            "2020-01-09 11:37:48 (31.3 MB/s) - ‘Labels_Valid.txt.4’ saved [14018/14018]\n",
            "\n",
            "Total Samples: 1147\n",
            "Total Training Samples: 918\n",
            "Total Validation Samples: 229\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "87Btnji1o76t",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "#Python generator to supply batches of traning data during training with loading full data set to memory\n",
        "def power_data_generator(batch_size,gen_type=''):\n",
        "    while 1:\n",
        "        df_input_train=pd.read_csv('./Input_Train.txt',usecols =['Wind_MWh','Actual_Load_MWh'],chunksize = 24*batch_size, iterator=True)\n",
        "        df_target_train=pd.read_csv('./Labels_Train.txt',usecols =['Potencia_Media_do_Vento_(MW)','Desvio_Padrao_Vento_(MW)','FCSPV_(Fracao_de_Carga_Suprida_pela_Potencia_do_Vento)'],chunksize = batch_size, iterator=True)\n",
        "        df_input_valid=pd.read_csv('./Input_Valid.txt',usecols =['Wind_MWh','Actual_Load_MWh'],chunksize = 24*batch_size, iterator=True)\n",
        "        df_target_valid=pd.read_csv('./Labels_Valid.txt',usecols =['Potencia_Media_do_Vento_(MW)','Desvio_Padrao_Vento_(MW)','FCSPV_(Fracao_de_Carga_Suprida_pela_Potencia_do_Vento)'],chunksize = batch_size, iterator=True)\n",
        "        for chunk, chunk2, chunk3, chunk4 in zip(df_input_train,df_target_train,df_input_valid,df_target_valid):\n",
        "          X_train = chunk.values\n",
        "          X_train = np.resize(X_train,(batch_size,24,2,1))\n",
        "          Y_train = chunk2.values\n",
        "          X_valid = chunk3.values\n",
        "          X_valid = np.resize(X_valid,(batch_size,24,2,1))\n",
        "          Y_valid = chunk4.values\n",
        "          if len(Y_train) < batch_size:\n",
        "            X_train = X_train[:-batch_size*len(Y_valid)]\n",
        "            Y_train = Y_train[:-len(Y_valid)]\n",
        "          if len(Y_valid) < batch_size:\n",
        "            X_valid = X_valid[:-batch_size*len(Y_valid)]\n",
        "            Y_valid = Y_valid[:-len(Y_valid)]       \n",
        "          if gen_type =='training':\n",
        "            yield (X_train[0:batch_size],Y_train[0:batch_size])\n",
        "          elif gen_type =='validation':\n",
        "            yield (X_valid[0:batch_size],Y_valid[0:batch_size])\n",
        "          elif gen_type =='inference':\n",
        "            yield X_valid"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "xib2u2mAlf0J",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "# Custom Metric\n",
        "\n",
        "def nossa_metrica(y_true, y_pred): # Overall metric\n",
        "  truepotventolow = K.cast(K.less_equal(y_true[:,0], 4000), 'int8')\n",
        "  predpotventolow = K.cast(K.less_equal(y_pred[:,0], 4000), 'int8')\n",
        "  potventolow = K.sum(truepotventolow*predpotventolow)\n",
        "  truepotventomed = K.cast(K.greater(y_true[:,0], 4000) & K.less_equal(y_true[:,0], 8500), 'int8')\n",
        "  predpotventomed = K.cast(K.greater(y_pred[:,0], 4000) & K.less_equal(y_pred[:,0], 8500), 'int8')\n",
        "  potventomed = K.sum(truepotventomed*predpotventomed)\n",
        "  truepotventohigh = K.cast(K.greater(y_true[:,0], 8500), 'int8')\n",
        "  predpotventohigh = K.cast(K.greater(y_pred[:,0], 8500), 'int8')\n",
        "  potventohigh = K.sum(truepotventohigh*predpotventohigh)\n",
        "  truedesvpadlow = K.cast(K.less_equal(y_true[:,1], 1150), 'int8')\n",
        "  preddesvpadlow = K.cast(K.less_equal(y_pred[:,1], 1150), 'int8')\n",
        "  desvpadlow = K.sum(truedesvpadlow*preddesvpadlow)\n",
        "  truedesvpadmed = K.cast(K.greater(y_true[:,1], 1150) & K.less_equal(y_true[:,1], 2300), 'int8')\n",
        "  preddesvpadmed = K.cast(K.greater(y_pred[:,1], 1150) & K.less_equal(y_pred[:,1], 2300), 'int8')\n",
        "  desvpadmed = K.sum(truedesvpadmed*preddesvpadmed)\n",
        "  truedesvpadhigh = K.cast(K.greater(y_true[:,1], 2300), 'int8')\n",
        "  preddesvpadhigh = K.cast(K.greater(y_pred[:,1], 2300), 'int8')\n",
        "  desvpadhigh = K.sum(truedesvpadhigh*preddesvpadhigh)\n",
        "  truewlslow = K.cast(K.less_equal(y_true[:,2], 0.075), 'int8')\n",
        "  predwlslow = K.cast(K.less_equal(y_pred[:,2], 0.075), 'int8')\n",
        "  wlslow = K.sum(truewlslow*predwlslow)   \n",
        "  truewlshigh = K.cast(K.greater(y_true[:,2], 0.075), 'int8')\n",
        "  predwlshigh = K.cast(K.greater(y_pred[:,2], 0.075), 'int8')\n",
        "  wlshigh = K.sum(truewlshigh*predwlshigh)\n",
        "  return (potventolow+potventomed+potventohigh+desvpadlow+desvpadmed+desvpadhigh+wlslow+wlshigh)/(Yclasses*batch_size)\n",
        "\n",
        "def potvento(y_true, y_pred): # Metric for first class\n",
        "  truepotventolow = K.cast(K.less_equal(y_true[:,0], 4000), 'int8')\n",
        "  predpotventolow = K.cast(K.less_equal(y_pred[:,0], 4000), 'int8')\n",
        "  potventolow = K.sum(truepotventolow*predpotventolow)\n",
        "  truepotventomed = K.cast(K.greater(y_true[:,0], 4000) & K.less_equal(y_true[:,0], 8500), 'int8')\n",
        "  predpotventomed = K.cast(K.greater(y_pred[:,0], 4000) & K.less_equal(y_pred[:,0], 8500), 'int8')\n",
        "  potventomed = K.sum(truepotventomed*predpotventomed)\n",
        "  truepotventohigh = K.cast(K.greater(y_true[:,0], 8500), 'int8')\n",
        "  predpotventohigh = K.cast(K.greater(y_pred[:,0], 8500), 'int8')\n",
        "  potventohigh = K.sum(truepotventohigh*predpotventohigh)\n",
        "  return (potventolow+potventomed+potventohigh)/(batch_size)\n",
        "\n",
        "def desvpad(y_true, y_pred): # Metric for second class\n",
        "  truedesvpadlow = K.cast(K.less_equal(y_true[:,1], 1150), 'int8')\n",
        "  preddesvpadlow = K.cast(K.less_equal(y_pred[:,1], 1150), 'int8')\n",
        "  desvpadlow = K.sum(truedesvpadlow*preddesvpadlow)\n",
        "  truedesvpadmed = K.cast(K.greater(y_true[:,1], 1150) & K.less_equal(y_true[:,1], 2300), 'int8')\n",
        "  preddesvpadmed = K.cast(K.greater(y_pred[:,1], 1150) & K.less_equal(y_pred[:,1], 2300), 'int8')\n",
        "  desvpadmed = K.sum(truedesvpadmed*preddesvpadmed)\n",
        "  truedesvpadhigh = K.cast(K.greater(y_true[:,1], 2300), 'int8')\n",
        "  preddesvpadhigh = K.cast(K.greater(y_pred[:,1], 2300), 'int8')\n",
        "  desvpadhigh = K.sum(truedesvpadhigh*preddesvpadhigh)    \n",
        "  return (desvpadlow+desvpadmed+desvpadhigh)/(batch_size)\n",
        "\n",
        "def wls(y_true, y_pred): # Metric for third class\n",
        "  truewlslow = K.cast(K.less_equal(y_true[:,2], 0.075), 'int8')\n",
        "  predwlslow = K.cast(K.less_equal(y_pred[:,2], 0.075), 'int8')\n",
        "  wlslow = K.sum(truewlslow*predwlslow)   \n",
        "  truewlshigh = K.cast(K.greater(y_true[:,2], 0.075), 'int8')\n",
        "  predwlshigh = K.cast(K.greater(y_pred[:,2], 0.075), 'int8')\n",
        "  wlshigh = K.sum(truewlshigh*predwlshigh)  \n",
        "  return (wlslow+wlshigh)/(batch_size)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "LTE-wn6gpG3F",
        "colab_type": "code",
        "outputId": "c9518737-4b48-40f8-a56f-151da6c93057",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 425
        }
      },
      "source": [
        "#Define model using Keras\n",
        "\n",
        "batch_size = 16\n",
        "Yclasses = 3 # Number of output classes\n",
        "bias0 =  'zeros'\n",
        "kernel0 = 'glorot_uniform'\n",
        "#dropout = 0.1\n",
        "\n",
        "\n",
        "model = keras.Sequential([\n",
        "    tf.keras.layers.Input(shape=(24,2,1),name='InputLayer'),                    \n",
        "    tf.keras.layers.Conv2D(filters=4,kernel_size=(6,2),strides=(1,1),activation='relu',bias_initializer=bias0,kernel_initializer=kernel0,name='ConvLayer1'),\n",
        "    #tf.keras.layers.Dropout(rate=dropout),\n",
        "    tf.keras.layers.Conv2D(filters=4,kernel_size=(6,1),strides=(1,1),activation='relu',bias_initializer=bias0,kernel_initializer=kernel0,name='ConvLayer2'),\n",
        "    #tf.keras.layers.Dropout(rate=dropout),\n",
        "    tf.keras.layers.Conv2D(filters=4,kernel_size=(5,1),strides=(1,1),activation='relu',bias_initializer=bias0,kernel_initializer=kernel0,name='ConvLayer3'),\n",
        "    #tf.keras.layers.Dropout(rate=dropout),\n",
        "    tf.keras.layers.Flatten(name=\"Flatten\"),\n",
        "    tf.keras.layers.Dense(units = 8,activation='relu',bias_initializer=bias0,kernel_initializer=kernel0,name='FeedForward1'),\n",
        "    #tf.keras.layers.Dropout(rate=dropout),\n",
        "    tf.keras.layers.Dense(units = 8,activation='relu',bias_initializer=bias0,kernel_initializer=kernel0,name='FeedForward2'),\n",
        "    #tf.keras.layers.Dropout(rate=dropout),\n",
        "    tf.keras.layers.Dense(units = 8,activation='relu',bias_initializer=bias0,kernel_initializer=kernel0,name='FeedForward3'),\n",
        "    #tf.keras.layers.Dropout(rate=dropout),\n",
        "    tf.keras.layers.Dense(units = Yclasses,bias_initializer=bias0,kernel_initializer=kernel0,name='OutputLayer'),\n",
        "])\n",
        "model.compile(loss='mse',optimizer='adam',metrics=[nossa_metrica,potvento,desvpad,wls])\n",
        "model.summary()"
      ],
      "execution_count": 5,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Model: \"sequential\"\n",
            "_________________________________________________________________\n",
            "Layer (type)                 Output Shape              Param #   \n",
            "=================================================================\n",
            "ConvLayer1 (Conv2D)          (None, 19, 1, 4)          52        \n",
            "_________________________________________________________________\n",
            "ConvLayer2 (Conv2D)          (None, 14, 1, 4)          100       \n",
            "_________________________________________________________________\n",
            "ConvLayer3 (Conv2D)          (None, 10, 1, 4)          84        \n",
            "_________________________________________________________________\n",
            "Flatten (Flatten)            (None, 40)                0         \n",
            "_________________________________________________________________\n",
            "FeedForward1 (Dense)         (None, 8)                 328       \n",
            "_________________________________________________________________\n",
            "FeedForward2 (Dense)         (None, 8)                 72        \n",
            "_________________________________________________________________\n",
            "FeedForward3 (Dense)         (None, 8)                 72        \n",
            "_________________________________________________________________\n",
            "OutputLayer (Dense)          (None, 3)                 27        \n",
            "=================================================================\n",
            "Total params: 735\n",
            "Trainable params: 735\n",
            "Non-trainable params: 0\n",
            "_________________________________________________________________\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "ToTJRHoV9Kdj",
        "colab_type": "code",
        "outputId": "87416913-38d0-49ac-c34e-5666e528f955",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1000
        }
      },
      "source": [
        "train_generator= power_data_generator(batch_size=batch_size,gen_type='training')\n",
        "valid_generator= power_data_generator(batch_size=batch_size,gen_type='validation')\n",
        "number_of_training_batches = np.int32(total_train/batch_size)\n",
        "number_of_validation_batches = np.int32(total_valid/batch_size) \n",
        "#Training starts\n",
        "t = time.time()\n",
        "history = model.fit_generator(train_generator, steps_per_epoch= number_of_training_batches,epochs=100,validation_data=valid_generator, validation_steps=number_of_validation_batches,verbose=2)\n",
        "print(\"Loss de treinamento minima:\",min(history.history[\"loss\"]))\n",
        "print(\"Acuracia de treinamento maxima:\",max(history.history[\"nossa_metrica\"]))\n",
        "print(\"Loss de valdiacao minima:\",min(history.history[\"val_loss\"]))\n",
        "print(\"Acuracia de validacao maxima:\",max(history.history[\"val_nossa_metrica\"])) \n",
        "print(\"Time: %.3f minutes\" % ((time.time() - t)/60))\n",
        "model.save('model.h5')\n",
        "print(\"Model Saved\")"
      ],
      "execution_count": 6,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "WARNING:tensorflow:From <ipython-input-6-5cc384f2e441>:7: Model.fit_generator (from tensorflow.python.keras.engine.training) is deprecated and will be removed in a future version.\n",
            "Instructions for updating:\n",
            "Please use Model.fit, which supports generators.\n",
            "WARNING:tensorflow:sample_weight modes were coerced from\n",
            "  ...\n",
            "    to  \n",
            "  ['...']\n",
            "WARNING:tensorflow:sample_weight modes were coerced from\n",
            "  ...\n",
            "    to  \n",
            "  ['...']\n",
            "Train for 57 steps, validate for 14 steps\n",
            "Epoch 1/100\n",
            "57/57 - 4s - loss: 7582500.9518 - nossa_metrica: 0.5570 - potvento: 0.5296 - desvpad: 0.5877 - wls: 0.5537 - val_loss: 14057465.2500 - val_nossa_metrica: 0.4167 - val_potvento: 0.3214 - val_desvpad: 0.4241 - val_wls: 0.5045\n",
            "Epoch 2/100\n",
            "57/57 - 1s - loss: 5411669.7259 - nossa_metrica: 0.5577 - potvento: 0.5395 - desvpad: 0.5680 - wls: 0.5658 - val_loss: 7857750.0357 - val_nossa_metrica: 0.3705 - val_potvento: 0.2679 - val_desvpad: 0.4018 - val_wls: 0.4420\n",
            "Epoch 3/100\n",
            "57/57 - 1s - loss: 2322246.4605 - nossa_metrica: 0.5325 - potvento: 0.5055 - desvpad: 0.5132 - wls: 0.5789 - val_loss: 3822801.9286 - val_nossa_metrica: 0.4122 - val_potvento: 0.3795 - val_desvpad: 0.4018 - val_wls: 0.4554\n",
            "Epoch 4/100\n",
            "57/57 - 1s - loss: 1608155.9320 - nossa_metrica: 0.5610 - potvento: 0.5011 - desvpad: 0.6053 - wls: 0.5768 - val_loss: 3212680.5446 - val_nossa_metrica: 0.4271 - val_potvento: 0.4152 - val_desvpad: 0.4062 - val_wls: 0.4598\n",
            "Epoch 5/100\n",
            "57/57 - 1s - loss: 1441048.2270 - nossa_metrica: 0.5779 - potvento: 0.5515 - desvpad: 0.6031 - wls: 0.5789 - val_loss: 2940616.4554 - val_nossa_metrica: 0.4420 - val_potvento: 0.4286 - val_desvpad: 0.4107 - val_wls: 0.4866\n",
            "Epoch 6/100\n",
            "57/57 - 1s - loss: 1381117.9539 - nossa_metrica: 0.5329 - potvento: 0.5548 - desvpad: 0.5833 - wls: 0.4605 - val_loss: 2414194.3661 - val_nossa_metrica: 0.3958 - val_potvento: 0.4330 - val_desvpad: 0.4062 - val_wls: 0.3482\n",
            "Epoch 7/100\n",
            "57/57 - 1s - loss: 1309814.3355 - nossa_metrica: 0.5358 - potvento: 0.5943 - desvpad: 0.5965 - wls: 0.4167 - val_loss: 2446822.6875 - val_nossa_metrica: 0.4152 - val_potvento: 0.4509 - val_desvpad: 0.4286 - val_wls: 0.3661\n",
            "Epoch 8/100\n",
            "57/57 - 1s - loss: 1230112.3092 - nossa_metrica: 0.5442 - potvento: 0.6404 - desvpad: 0.6009 - wls: 0.3914 - val_loss: 2230929.7589 - val_nossa_metrica: 0.4345 - val_potvento: 0.4464 - val_desvpad: 0.4196 - val_wls: 0.4375\n",
            "Epoch 9/100\n",
            "57/57 - 1s - loss: 1149106.1535 - nossa_metrica: 0.5475 - potvento: 0.6568 - desvpad: 0.6020 - wls: 0.3838 - val_loss: 2289156.4821 - val_nossa_metrica: 0.4449 - val_potvento: 0.4732 - val_desvpad: 0.4062 - val_wls: 0.4554\n",
            "Epoch 10/100\n",
            "57/57 - 1s - loss: 1092959.7971 - nossa_metrica: 0.5665 - potvento: 0.6952 - desvpad: 0.6075 - wls: 0.3969 - val_loss: 2063101.5179 - val_nossa_metrica: 0.4271 - val_potvento: 0.4464 - val_desvpad: 0.4330 - val_wls: 0.4018\n",
            "Epoch 11/100\n",
            "57/57 - 1s - loss: 1014721.5724 - nossa_metrica: 0.5760 - potvento: 0.7149 - desvpad: 0.5910 - wls: 0.4221 - val_loss: 1888399.1518 - val_nossa_metrica: 0.4256 - val_potvento: 0.4152 - val_desvpad: 0.4464 - val_wls: 0.4152\n",
            "Epoch 12/100\n",
            "57/57 - 1s - loss: 970332.2357 - nossa_metrica: 0.5928 - potvento: 0.7533 - desvpad: 0.6053 - wls: 0.4200 - val_loss: 1730075.2098 - val_nossa_metrica: 0.4390 - val_potvento: 0.4598 - val_desvpad: 0.4464 - val_wls: 0.4107\n",
            "Epoch 13/100\n",
            "57/57 - 1s - loss: 892484.7829 - nossa_metrica: 0.5965 - potvento: 0.7763 - desvpad: 0.5965 - wls: 0.4167 - val_loss: 1595627.5000 - val_nossa_metrica: 0.4539 - val_potvento: 0.4777 - val_desvpad: 0.4732 - val_wls: 0.4107\n",
            "Epoch 14/100\n",
            "57/57 - 1s - loss: 793380.3794 - nossa_metrica: 0.6009 - potvento: 0.7993 - desvpad: 0.5833 - wls: 0.4200 - val_loss: 1448259.0402 - val_nossa_metrica: 0.4702 - val_potvento: 0.4911 - val_desvpad: 0.5045 - val_wls: 0.4152\n",
            "Epoch 15/100\n",
            "57/57 - 1s - loss: 692047.6096 - nossa_metrica: 0.6151 - potvento: 0.8366 - desvpad: 0.5713 - wls: 0.4375 - val_loss: 1158962.7723 - val_nossa_metrica: 0.4836 - val_potvento: 0.5045 - val_desvpad: 0.4955 - val_wls: 0.4509\n",
            "Epoch 16/100\n",
            "57/57 - 1s - loss: 539503.9446 - nossa_metrica: 0.6396 - potvento: 0.8684 - desvpad: 0.5581 - wls: 0.4923 - val_loss: 897891.7991 - val_nossa_metrica: 0.5625 - val_potvento: 0.6518 - val_desvpad: 0.5491 - val_wls: 0.4866\n",
            "Epoch 17/100\n",
            "57/57 - 1s - loss: 379390.5249 - nossa_metrica: 0.6612 - potvento: 0.9057 - desvpad: 0.5625 - wls: 0.5154 - val_loss: 522517.8661 - val_nossa_metrica: 0.5908 - val_potvento: 0.7366 - val_desvpad: 0.4955 - val_wls: 0.5402\n",
            "Epoch 18/100\n",
            "57/57 - 1s - loss: 226149.9227 - nossa_metrica: 0.6985 - potvento: 0.9353 - desvpad: 0.5691 - wls: 0.5910 - val_loss: 346814.4129 - val_nossa_metrica: 0.6220 - val_potvento: 0.8080 - val_desvpad: 0.4732 - val_wls: 0.5848\n",
            "Epoch 19/100\n",
            "57/57 - 1s - loss: 165407.7119 - nossa_metrica: 0.7156 - potvento: 0.9364 - desvpad: 0.5647 - wls: 0.6458 - val_loss: 288486.3705 - val_nossa_metrica: 0.6429 - val_potvento: 0.8348 - val_desvpad: 0.4911 - val_wls: 0.6027\n",
            "Epoch 20/100\n",
            "57/57 - 1s - loss: 145566.9062 - nossa_metrica: 0.7387 - potvento: 0.9529 - desvpad: 0.5724 - wls: 0.6908 - val_loss: 265482.0536 - val_nossa_metrica: 0.6369 - val_potvento: 0.8259 - val_desvpad: 0.5000 - val_wls: 0.5848\n",
            "Epoch 21/100\n",
            "57/57 - 1s - loss: 133633.6369 - nossa_metrica: 0.7240 - potvento: 0.9561 - desvpad: 0.5603 - wls: 0.6557 - val_loss: 226112.0419 - val_nossa_metrica: 0.6592 - val_potvento: 0.8527 - val_desvpad: 0.5312 - val_wls: 0.5938\n",
            "Epoch 22/100\n",
            "57/57 - 1s - loss: 117532.1730 - nossa_metrica: 0.7310 - potvento: 0.9627 - desvpad: 0.5954 - wls: 0.6349 - val_loss: 219334.1864 - val_nossa_metrica: 0.6577 - val_potvento: 0.8661 - val_desvpad: 0.5179 - val_wls: 0.5893\n",
            "Epoch 23/100\n",
            "57/57 - 1s - loss: 107449.7767 - nossa_metrica: 0.7394 - potvento: 0.9715 - desvpad: 0.6217 - wls: 0.6250 - val_loss: 200402.8175 - val_nossa_metrica: 0.6667 - val_potvento: 0.8571 - val_desvpad: 0.5446 - val_wls: 0.5982\n",
            "Epoch 24/100\n",
            "57/57 - 1s - loss: 98685.7229 - nossa_metrica: 0.7474 - potvento: 0.9737 - desvpad: 0.6349 - wls: 0.6338 - val_loss: 203424.2461 - val_nossa_metrica: 0.6607 - val_potvento: 0.8571 - val_desvpad: 0.5268 - val_wls: 0.5982\n",
            "Epoch 25/100\n",
            "57/57 - 1s - loss: 90715.3949 - nossa_metrica: 0.7548 - potvento: 0.9770 - desvpad: 0.6502 - wls: 0.6371 - val_loss: 190064.7243 - val_nossa_metrica: 0.6577 - val_potvento: 0.8616 - val_desvpad: 0.5312 - val_wls: 0.5804\n",
            "Epoch 26/100\n",
            "57/57 - 1s - loss: 86292.4893 - nossa_metrica: 0.7569 - potvento: 0.9704 - desvpad: 0.6404 - wls: 0.6601 - val_loss: 167415.9927 - val_nossa_metrica: 0.6622 - val_potvento: 0.8616 - val_desvpad: 0.5446 - val_wls: 0.5804\n",
            "Epoch 27/100\n",
            "57/57 - 1s - loss: 79115.7257 - nossa_metrica: 0.7628 - potvento: 0.9704 - desvpad: 0.6678 - wls: 0.6502 - val_loss: 158749.1649 - val_nossa_metrica: 0.6592 - val_potvento: 0.8527 - val_desvpad: 0.5491 - val_wls: 0.5759\n",
            "Epoch 28/100\n",
            "57/57 - 1s - loss: 73368.7601 - nossa_metrica: 0.7683 - potvento: 0.9726 - desvpad: 0.6919 - wls: 0.6404 - val_loss: 145696.4562 - val_nossa_metrica: 0.6652 - val_potvento: 0.8571 - val_desvpad: 0.5670 - val_wls: 0.5714\n",
            "Epoch 29/100\n",
            "57/57 - 1s - loss: 68373.1614 - nossa_metrica: 0.7792 - potvento: 0.9737 - desvpad: 0.7193 - wls: 0.6447 - val_loss: 146755.2369 - val_nossa_metrica: 0.6771 - val_potvento: 0.8750 - val_desvpad: 0.5804 - val_wls: 0.5759\n",
            "Epoch 30/100\n",
            "57/57 - 1s - loss: 63241.0942 - nossa_metrica: 0.7855 - potvento: 0.9759 - desvpad: 0.7248 - wls: 0.6557 - val_loss: 126508.7874 - val_nossa_metrica: 0.6786 - val_potvento: 0.8661 - val_desvpad: 0.5938 - val_wls: 0.5759\n",
            "Epoch 31/100\n",
            "57/57 - 1s - loss: 60814.1168 - nossa_metrica: 0.7876 - potvento: 0.9748 - desvpad: 0.7292 - wls: 0.6590 - val_loss: 128103.7907 - val_nossa_metrica: 0.7336 - val_potvento: 0.9330 - val_desvpad: 0.6562 - val_wls: 0.6116\n",
            "Epoch 32/100\n",
            "57/57 - 1s - loss: 55466.9113 - nossa_metrica: 0.7961 - potvento: 0.9748 - desvpad: 0.7588 - wls: 0.6546 - val_loss: 119642.0706 - val_nossa_metrica: 0.6786 - val_potvento: 0.8705 - val_desvpad: 0.6027 - val_wls: 0.5625\n",
            "Epoch 33/100\n",
            "57/57 - 1s - loss: 52286.9414 - nossa_metrica: 0.7982 - potvento: 0.9814 - desvpad: 0.7719 - wls: 0.6414 - val_loss: 112511.6685 - val_nossa_metrica: 0.6845 - val_potvento: 0.8705 - val_desvpad: 0.6116 - val_wls: 0.5714\n",
            "Epoch 34/100\n",
            "57/57 - 1s - loss: 48918.3328 - nossa_metrica: 0.8023 - potvento: 0.9825 - desvpad: 0.7774 - wls: 0.6469 - val_loss: 105421.0692 - val_nossa_metrica: 0.6979 - val_potvento: 0.8839 - val_desvpad: 0.6384 - val_wls: 0.5714\n",
            "Epoch 35/100\n",
            "57/57 - 1s - loss: 45049.5436 - nossa_metrica: 0.8059 - potvento: 0.9846 - desvpad: 0.7862 - wls: 0.6469 - val_loss: 95531.2776 - val_nossa_metrica: 0.7083 - val_potvento: 0.8795 - val_desvpad: 0.6607 - val_wls: 0.5848\n",
            "Epoch 36/100\n",
            "57/57 - 1s - loss: 43756.8412 - nossa_metrica: 0.8059 - potvento: 0.9803 - desvpad: 0.7873 - wls: 0.6502 - val_loss: 78696.1632 - val_nossa_metrica: 0.7173 - val_potvento: 0.8795 - val_desvpad: 0.6830 - val_wls: 0.5893\n",
            "Epoch 37/100\n",
            "57/57 - 1s - loss: 39540.3196 - nossa_metrica: 0.8085 - potvento: 0.9857 - desvpad: 0.8114 - wls: 0.6283 - val_loss: 79673.8114 - val_nossa_metrica: 0.7188 - val_potvento: 0.8973 - val_desvpad: 0.6652 - val_wls: 0.5938\n",
            "Epoch 38/100\n",
            "57/57 - 1s - loss: 37188.3381 - nossa_metrica: 0.8110 - potvento: 0.9890 - desvpad: 0.8136 - wls: 0.6305 - val_loss: 73031.1754 - val_nossa_metrica: 0.7307 - val_potvento: 0.8973 - val_desvpad: 0.7054 - val_wls: 0.5893\n",
            "Epoch 39/100\n",
            "57/57 - 1s - loss: 34658.3164 - nossa_metrica: 0.8220 - potvento: 0.9868 - desvpad: 0.8224 - wls: 0.6568 - val_loss: 75344.6378 - val_nossa_metrica: 0.7232 - val_potvento: 0.8929 - val_desvpad: 0.7009 - val_wls: 0.5759\n",
            "Epoch 40/100\n",
            "57/57 - 1s - loss: 32019.8917 - nossa_metrica: 0.8308 - potvento: 0.9879 - desvpad: 0.8366 - wls: 0.6678 - val_loss: 69446.9051 - val_nossa_metrica: 0.7351 - val_potvento: 0.8839 - val_desvpad: 0.7277 - val_wls: 0.5938\n",
            "Epoch 41/100\n",
            "57/57 - 1s - loss: 31593.3415 - nossa_metrica: 0.8242 - potvento: 0.9868 - desvpad: 0.8202 - wls: 0.6656 - val_loss: 62752.4297 - val_nossa_metrica: 0.7321 - val_potvento: 0.8884 - val_desvpad: 0.6964 - val_wls: 0.6116\n",
            "Epoch 42/100\n",
            "57/57 - 1s - loss: 29594.7423 - nossa_metrica: 0.8257 - potvento: 0.9825 - desvpad: 0.8322 - wls: 0.6623 - val_loss: 61137.4276 - val_nossa_metrica: 0.7351 - val_potvento: 0.9062 - val_desvpad: 0.6920 - val_wls: 0.6071\n",
            "Epoch 43/100\n",
            "57/57 - 1s - loss: 28697.5984 - nossa_metrica: 0.8311 - potvento: 0.9846 - desvpad: 0.8366 - wls: 0.6721 - val_loss: 59575.9869 - val_nossa_metrica: 0.7396 - val_potvento: 0.9018 - val_desvpad: 0.7232 - val_wls: 0.5938\n",
            "Epoch 44/100\n",
            "57/57 - 1s - loss: 28780.8046 - nossa_metrica: 0.8344 - potvento: 0.9792 - desvpad: 0.8322 - wls: 0.6919 - val_loss: 56121.6415 - val_nossa_metrica: 0.7485 - val_potvento: 0.9062 - val_desvpad: 0.7321 - val_wls: 0.6071\n",
            "Epoch 45/100\n",
            "57/57 - 1s - loss: 29403.4203 - nossa_metrica: 0.8395 - potvento: 0.9792 - desvpad: 0.8377 - wls: 0.7018 - val_loss: 90532.1818 - val_nossa_metrica: 0.7217 - val_potvento: 0.8661 - val_desvpad: 0.6964 - val_wls: 0.6027\n",
            "Epoch 46/100\n",
            "57/57 - 1s - loss: 31793.4960 - nossa_metrica: 0.8330 - potvento: 0.9770 - desvpad: 0.8279 - wls: 0.6941 - val_loss: 86935.2450 - val_nossa_metrica: 0.7857 - val_potvento: 0.9375 - val_desvpad: 0.7723 - val_wls: 0.6473\n",
            "Epoch 47/100\n",
            "57/57 - 1s - loss: 33364.3475 - nossa_metrica: 0.8410 - potvento: 0.9693 - desvpad: 0.8333 - wls: 0.7204 - val_loss: 61671.1823 - val_nossa_metrica: 0.7485 - val_potvento: 0.8973 - val_desvpad: 0.7321 - val_wls: 0.6161\n",
            "Epoch 48/100\n",
            "57/57 - 1s - loss: 31593.4635 - nossa_metrica: 0.8439 - potvento: 0.9682 - desvpad: 0.8355 - wls: 0.7281 - val_loss: 78165.4665 - val_nossa_metrica: 0.7396 - val_potvento: 0.8661 - val_desvpad: 0.7321 - val_wls: 0.6205\n",
            "Epoch 49/100\n",
            "57/57 - 1s - loss: 30054.8216 - nossa_metrica: 0.8476 - potvento: 0.9649 - desvpad: 0.8399 - wls: 0.7379 - val_loss: 48516.5706 - val_nossa_metrica: 0.7485 - val_potvento: 0.9062 - val_desvpad: 0.7321 - val_wls: 0.6071\n",
            "Epoch 50/100\n",
            "57/57 - 1s - loss: 27692.4794 - nossa_metrica: 0.8498 - potvento: 0.9682 - desvpad: 0.8465 - wls: 0.7346 - val_loss: 77054.2573 - val_nossa_metrica: 0.7307 - val_potvento: 0.8705 - val_desvpad: 0.7098 - val_wls: 0.6116\n",
            "Epoch 51/100\n",
            "57/57 - 1s - loss: 28063.6973 - nossa_metrica: 0.8447 - potvento: 0.9682 - desvpad: 0.8399 - wls: 0.7259 - val_loss: 50978.5756 - val_nossa_metrica: 0.7411 - val_potvento: 0.8795 - val_desvpad: 0.7545 - val_wls: 0.5893\n",
            "Epoch 52/100\n",
            "57/57 - 1s - loss: 26637.5750 - nossa_metrica: 0.8490 - potvento: 0.9704 - desvpad: 0.8454 - wls: 0.7314 - val_loss: 54601.5723 - val_nossa_metrica: 0.7485 - val_potvento: 0.8929 - val_desvpad: 0.7277 - val_wls: 0.6250\n",
            "Epoch 53/100\n",
            "57/57 - 1s - loss: 25607.4628 - nossa_metrica: 0.8509 - potvento: 0.9748 - desvpad: 0.8520 - wls: 0.7259 - val_loss: 59873.6575 - val_nossa_metrica: 0.7485 - val_potvento: 0.8839 - val_desvpad: 0.7545 - val_wls: 0.6071\n",
            "Epoch 54/100\n",
            "57/57 - 1s - loss: 24991.1483 - nossa_metrica: 0.8534 - potvento: 0.9726 - desvpad: 0.8487 - wls: 0.7390 - val_loss: 47024.6652 - val_nossa_metrica: 0.7366 - val_potvento: 0.9107 - val_desvpad: 0.7232 - val_wls: 0.5759\n",
            "Epoch 55/100\n",
            "57/57 - 1s - loss: 23768.2082 - nossa_metrica: 0.8567 - potvento: 0.9770 - desvpad: 0.8542 - wls: 0.7390 - val_loss: 67715.8838 - val_nossa_metrica: 0.7277 - val_potvento: 0.8661 - val_desvpad: 0.7366 - val_wls: 0.5804\n",
            "Epoch 56/100\n",
            "57/57 - 1s - loss: 24541.6230 - nossa_metrica: 0.8472 - potvento: 0.9792 - desvpad: 0.8443 - wls: 0.7182 - val_loss: 51459.6141 - val_nossa_metrica: 0.7411 - val_potvento: 0.8884 - val_desvpad: 0.7277 - val_wls: 0.6071\n",
            "Epoch 57/100\n",
            "57/57 - 1s - loss: 23857.6061 - nossa_metrica: 0.8516 - potvento: 0.9770 - desvpad: 0.8542 - wls: 0.7237 - val_loss: 48180.6892 - val_nossa_metrica: 0.7426 - val_potvento: 0.9018 - val_desvpad: 0.7321 - val_wls: 0.5938\n",
            "Epoch 58/100\n",
            "57/57 - 1s - loss: 23475.9245 - nossa_metrica: 0.8545 - potvento: 0.9781 - desvpad: 0.8553 - wls: 0.7303 - val_loss: 58360.3259 - val_nossa_metrica: 0.7366 - val_potvento: 0.8884 - val_desvpad: 0.7366 - val_wls: 0.5848\n",
            "Epoch 59/100\n",
            "57/57 - 1s - loss: 23431.0353 - nossa_metrica: 0.8527 - potvento: 0.9759 - desvpad: 0.8520 - wls: 0.7303 - val_loss: 43011.8684 - val_nossa_metrica: 0.7530 - val_potvento: 0.9107 - val_desvpad: 0.7500 - val_wls: 0.5982\n",
            "Epoch 60/100\n",
            "57/57 - 1s - loss: 22299.6484 - nossa_metrica: 0.8534 - potvento: 0.9748 - desvpad: 0.8553 - wls: 0.7303 - val_loss: 62702.8969 - val_nossa_metrica: 0.7217 - val_potvento: 0.8705 - val_desvpad: 0.7277 - val_wls: 0.5670\n",
            "Epoch 61/100\n",
            "57/57 - 1s - loss: 22954.8618 - nossa_metrica: 0.8439 - potvento: 0.9748 - desvpad: 0.8476 - wls: 0.7094 - val_loss: 57952.5081 - val_nossa_metrica: 0.7798 - val_potvento: 0.9464 - val_desvpad: 0.7946 - val_wls: 0.5982\n",
            "Epoch 62/100\n",
            "57/57 - 1s - loss: 22261.4136 - nossa_metrica: 0.8406 - potvento: 0.9759 - desvpad: 0.8575 - wls: 0.6886 - val_loss: 46206.0040 - val_nossa_metrica: 0.7426 - val_potvento: 0.9062 - val_desvpad: 0.7455 - val_wls: 0.5759\n",
            "Epoch 63/100\n",
            "57/57 - 1s - loss: 21972.8026 - nossa_metrica: 0.8421 - potvento: 0.9759 - desvpad: 0.8596 - wls: 0.6908 - val_loss: 57316.4224 - val_nossa_metrica: 0.7232 - val_potvento: 0.8884 - val_desvpad: 0.7232 - val_wls: 0.5580\n",
            "Epoch 64/100\n",
            "57/57 - 1s - loss: 21907.2731 - nossa_metrica: 0.8428 - potvento: 0.9737 - desvpad: 0.8607 - wls: 0.6941 - val_loss: 44155.8820 - val_nossa_metrica: 0.7366 - val_potvento: 0.9107 - val_desvpad: 0.7500 - val_wls: 0.5491\n",
            "Epoch 65/100\n",
            "57/57 - 1s - loss: 20335.6486 - nossa_metrica: 0.8417 - potvento: 0.9748 - desvpad: 0.8684 - wls: 0.6820 - val_loss: 54146.0836 - val_nossa_metrica: 0.7068 - val_potvento: 0.8750 - val_desvpad: 0.7277 - val_wls: 0.5179\n",
            "Epoch 66/100\n",
            "57/57 - 1s - loss: 20717.2857 - nossa_metrica: 0.8330 - potvento: 0.9748 - desvpad: 0.8575 - wls: 0.6667 - val_loss: 46932.4693 - val_nossa_metrica: 0.7054 - val_potvento: 0.8795 - val_desvpad: 0.7500 - val_wls: 0.4866\n",
            "Epoch 67/100\n",
            "57/57 - 1s - loss: 19973.7046 - nossa_metrica: 0.8355 - potvento: 0.9759 - desvpad: 0.8629 - wls: 0.6678 - val_loss: 42049.7922 - val_nossa_metrica: 0.7158 - val_potvento: 0.9107 - val_desvpad: 0.7277 - val_wls: 0.5089\n",
            "Epoch 68/100\n",
            "57/57 - 1s - loss: 19675.1054 - nossa_metrica: 0.8304 - potvento: 0.9759 - desvpad: 0.8662 - wls: 0.6491 - val_loss: 46756.3778 - val_nossa_metrica: 0.7113 - val_potvento: 0.8929 - val_desvpad: 0.7545 - val_wls: 0.4866\n",
            "Epoch 69/100\n",
            "57/57 - 1s - loss: 19720.7471 - nossa_metrica: 0.8249 - potvento: 0.9737 - desvpad: 0.8662 - wls: 0.6349 - val_loss: 44140.2165 - val_nossa_metrica: 0.7039 - val_potvento: 0.9107 - val_desvpad: 0.7366 - val_wls: 0.4643\n",
            "Epoch 70/100\n",
            "57/57 - 1s - loss: 18981.3334 - nossa_metrica: 0.8213 - potvento: 0.9781 - desvpad: 0.8673 - wls: 0.6184 - val_loss: 50733.7423 - val_nossa_metrica: 0.6979 - val_potvento: 0.8929 - val_desvpad: 0.7545 - val_wls: 0.4464\n",
            "Epoch 71/100\n",
            "57/57 - 1s - loss: 19734.8005 - nossa_metrica: 0.8074 - potvento: 0.9825 - desvpad: 0.8509 - wls: 0.5888 - val_loss: 46071.2147 - val_nossa_metrica: 0.7039 - val_potvento: 0.8884 - val_desvpad: 0.7366 - val_wls: 0.4866\n",
            "Epoch 72/100\n",
            "57/57 - 1s - loss: 19068.7214 - nossa_metrica: 0.8114 - potvento: 0.9825 - desvpad: 0.8564 - wls: 0.5954 - val_loss: 40711.3709 - val_nossa_metrica: 0.7173 - val_potvento: 0.9196 - val_desvpad: 0.7232 - val_wls: 0.5089\n",
            "Epoch 73/100\n",
            "57/57 - 1s - loss: 18698.6857 - nossa_metrica: 0.8114 - potvento: 0.9770 - desvpad: 0.8629 - wls: 0.5943 - val_loss: 45113.6523 - val_nossa_metrica: 0.7128 - val_potvento: 0.9018 - val_desvpad: 0.7411 - val_wls: 0.4955\n",
            "Epoch 74/100\n",
            "57/57 - 1s - loss: 19127.8594 - nossa_metrica: 0.8026 - potvento: 0.9857 - desvpad: 0.8520 - wls: 0.5702 - val_loss: 40038.4009 - val_nossa_metrica: 0.7098 - val_potvento: 0.9107 - val_desvpad: 0.7545 - val_wls: 0.4643\n",
            "Epoch 75/100\n",
            "57/57 - 1s - loss: 18263.2092 - nossa_metrica: 0.8041 - potvento: 0.9836 - desvpad: 0.8575 - wls: 0.5713 - val_loss: 45238.4872 - val_nossa_metrica: 0.6979 - val_potvento: 0.8929 - val_desvpad: 0.7366 - val_wls: 0.4643\n",
            "Epoch 76/100\n",
            "57/57 - 1s - loss: 18705.8897 - nossa_metrica: 0.7975 - potvento: 0.9857 - desvpad: 0.8531 - wls: 0.5537 - val_loss: 50715.9595 - val_nossa_metrica: 0.7530 - val_potvento: 0.9643 - val_desvpad: 0.7902 - val_wls: 0.5045\n",
            "Epoch 77/100\n",
            "57/57 - 1s - loss: 18142.2704 - nossa_metrica: 0.7986 - potvento: 0.9868 - desvpad: 0.8564 - wls: 0.5526 - val_loss: 41351.6250 - val_nossa_metrica: 0.7039 - val_potvento: 0.9107 - val_desvpad: 0.7455 - val_wls: 0.4554\n",
            "Epoch 78/100\n",
            "57/57 - 1s - loss: 18065.0574 - nossa_metrica: 0.7968 - potvento: 0.9890 - desvpad: 0.8553 - wls: 0.5461 - val_loss: 42795.3426 - val_nossa_metrica: 0.7009 - val_potvento: 0.9107 - val_desvpad: 0.7366 - val_wls: 0.4554\n",
            "Epoch 79/100\n",
            "57/57 - 1s - loss: 18114.7031 - nossa_metrica: 0.7935 - potvento: 0.9846 - desvpad: 0.8542 - wls: 0.5417 - val_loss: 41749.1996 - val_nossa_metrica: 0.6979 - val_potvento: 0.9018 - val_desvpad: 0.7500 - val_wls: 0.4420\n",
            "Epoch 80/100\n",
            "57/57 - 1s - loss: 17305.7949 - nossa_metrica: 0.7993 - potvento: 0.9890 - desvpad: 0.8629 - wls: 0.5461 - val_loss: 42097.9009 - val_nossa_metrica: 0.6949 - val_potvento: 0.9018 - val_desvpad: 0.7455 - val_wls: 0.4375\n",
            "Epoch 81/100\n",
            "57/57 - 1s - loss: 17815.8610 - nossa_metrica: 0.7931 - potvento: 0.9901 - desvpad: 0.8586 - wls: 0.5307 - val_loss: 38997.5836 - val_nossa_metrica: 0.6994 - val_potvento: 0.8929 - val_desvpad: 0.7545 - val_wls: 0.4509\n",
            "Epoch 82/100\n",
            "57/57 - 1s - loss: 17267.8412 - nossa_metrica: 0.8015 - potvento: 0.9890 - desvpad: 0.8629 - wls: 0.5526 - val_loss: 39684.9960 - val_nossa_metrica: 0.6964 - val_potvento: 0.9152 - val_desvpad: 0.7321 - val_wls: 0.4420\n",
            "Epoch 83/100\n",
            "57/57 - 1s - loss: 16997.7549 - nossa_metrica: 0.8023 - potvento: 0.9868 - desvpad: 0.8662 - wls: 0.5537 - val_loss: 39204.6762 - val_nossa_metrica: 0.6935 - val_potvento: 0.9152 - val_desvpad: 0.7455 - val_wls: 0.4196\n",
            "Epoch 84/100\n",
            "57/57 - 1s - loss: 17147.3472 - nossa_metrica: 0.8037 - potvento: 0.9868 - desvpad: 0.8684 - wls: 0.5559 - val_loss: 39935.7342 - val_nossa_metrica: 0.6920 - val_potvento: 0.9152 - val_desvpad: 0.7455 - val_wls: 0.4152\n",
            "Epoch 85/100\n",
            "57/57 - 1s - loss: 16335.6077 - nossa_metrica: 0.8077 - potvento: 0.9879 - desvpad: 0.8739 - wls: 0.5614 - val_loss: 41304.6014 - val_nossa_metrica: 0.7039 - val_potvento: 0.9062 - val_desvpad: 0.7634 - val_wls: 0.4420\n",
            "Epoch 86/100\n",
            "57/57 - 1s - loss: 16790.4936 - nossa_metrica: 0.7993 - potvento: 0.9879 - desvpad: 0.8684 - wls: 0.5417 - val_loss: 38175.7547 - val_nossa_metrica: 0.6949 - val_potvento: 0.8973 - val_desvpad: 0.7455 - val_wls: 0.4420\n",
            "Epoch 87/100\n",
            "57/57 - 1s - loss: 16291.4743 - nossa_metrica: 0.8056 - potvento: 0.9879 - desvpad: 0.8728 - wls: 0.5559 - val_loss: 37489.5737 - val_nossa_metrica: 0.7024 - val_potvento: 0.9152 - val_desvpad: 0.7366 - val_wls: 0.4554\n",
            "Epoch 88/100\n",
            "57/57 - 1s - loss: 16130.5296 - nossa_metrica: 0.8074 - potvento: 0.9868 - desvpad: 0.8739 - wls: 0.5614 - val_loss: 39181.1384 - val_nossa_metrica: 0.6994 - val_potvento: 0.9107 - val_desvpad: 0.7321 - val_wls: 0.4554\n",
            "Epoch 89/100\n",
            "57/57 - 1s - loss: 16264.1252 - nossa_metrica: 0.8037 - potvento: 0.9868 - desvpad: 0.8684 - wls: 0.5559 - val_loss: 35030.7142 - val_nossa_metrica: 0.7054 - val_potvento: 0.9107 - val_desvpad: 0.7679 - val_wls: 0.4375\n",
            "Epoch 90/100\n",
            "57/57 - 1s - loss: 15470.1146 - nossa_metrica: 0.8081 - potvento: 0.9879 - desvpad: 0.8750 - wls: 0.5614 - val_loss: 38558.0190 - val_nossa_metrica: 0.7054 - val_potvento: 0.9062 - val_desvpad: 0.7500 - val_wls: 0.4598\n",
            "Epoch 91/100\n",
            "57/57 - 1s - loss: 15872.7586 - nossa_metrica: 0.8012 - potvento: 0.9879 - desvpad: 0.8750 - wls: 0.5406 - val_loss: 41107.5086 - val_nossa_metrica: 0.7515 - val_potvento: 0.9688 - val_desvpad: 0.8125 - val_wls: 0.4732\n",
            "Epoch 92/100\n",
            "57/57 - 1s - loss: 15381.1814 - nossa_metrica: 0.8045 - potvento: 0.9879 - desvpad: 0.8816 - wls: 0.5439 - val_loss: 36817.5033 - val_nossa_metrica: 0.7054 - val_potvento: 0.9107 - val_desvpad: 0.7545 - val_wls: 0.4509\n",
            "Epoch 93/100\n",
            "57/57 - 1s - loss: 15267.4603 - nossa_metrica: 0.8034 - potvento: 0.9868 - desvpad: 0.8827 - wls: 0.5406 - val_loss: 38671.8659 - val_nossa_metrica: 0.6964 - val_potvento: 0.9107 - val_desvpad: 0.7366 - val_wls: 0.4420\n",
            "Epoch 94/100\n",
            "57/57 - 1s - loss: 15506.9756 - nossa_metrica: 0.7975 - potvento: 0.9868 - desvpad: 0.8750 - wls: 0.5307 - val_loss: 36371.4337 - val_nossa_metrica: 0.7024 - val_potvento: 0.9062 - val_desvpad: 0.7634 - val_wls: 0.4375\n",
            "Epoch 95/100\n",
            "57/57 - 1s - loss: 14831.3928 - nossa_metrica: 0.7990 - potvento: 0.9879 - desvpad: 0.8805 - wls: 0.5285 - val_loss: 36835.1662 - val_nossa_metrica: 0.7039 - val_potvento: 0.9062 - val_desvpad: 0.7545 - val_wls: 0.4509\n",
            "Epoch 96/100\n",
            "57/57 - 1s - loss: 15457.8951 - nossa_metrica: 0.7855 - potvento: 0.9879 - desvpad: 0.8728 - wls: 0.4956 - val_loss: 34680.6656 - val_nossa_metrica: 0.6905 - val_potvento: 0.8973 - val_desvpad: 0.7545 - val_wls: 0.4196\n",
            "Epoch 97/100\n",
            "57/57 - 1s - loss: 15071.8125 - nossa_metrica: 0.7891 - potvento: 0.9879 - desvpad: 0.8750 - wls: 0.5044 - val_loss: 34889.5098 - val_nossa_metrica: 0.7024 - val_potvento: 0.9062 - val_desvpad: 0.7455 - val_wls: 0.4554\n",
            "Epoch 98/100\n",
            "57/57 - 1s - loss: 14995.4513 - nossa_metrica: 0.7931 - potvento: 0.9868 - desvpad: 0.8794 - wls: 0.5132 - val_loss: 36698.5043 - val_nossa_metrica: 0.7024 - val_potvento: 0.9062 - val_desvpad: 0.7723 - val_wls: 0.4286\n",
            "Epoch 99/100\n",
            "57/57 - 1s - loss: 15185.9028 - nossa_metrica: 0.7909 - potvento: 0.9879 - desvpad: 0.8750 - wls: 0.5099 - val_loss: 36281.7210 - val_nossa_metrica: 0.7039 - val_potvento: 0.9107 - val_desvpad: 0.7679 - val_wls: 0.4330\n",
            "Epoch 100/100\n",
            "57/57 - 1s - loss: 14442.8978 - nossa_metrica: 0.7957 - potvento: 0.9879 - desvpad: 0.8827 - wls: 0.5164 - val_loss: 38006.5166 - val_nossa_metrica: 0.6920 - val_potvento: 0.9062 - val_desvpad: 0.7589 - val_wls: 0.4107\n",
            "Loss de treinamento minima: 14442.89776075932\n",
            "Acuracia de treinamento maxima: 0.8567252\n",
            "Loss de valdiacao minima: 34680.66559709822\n",
            "Acuracia de validacao maxima: 0.78571427\n",
            "Time: 1.344 minutes\n",
            "Model Saved\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "1hRaMx9CEyhS",
        "colab_type": "code",
        "outputId": "605cfe42-e3aa-4533-a697-84eb564cba6e",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1000
        }
      },
      "source": [
        "inference_generator= power_data_generator(batch_size=1,gen_type='inference')  #Generator for getting only input data\n",
        "test_input = next(inference_generator)  #Get next batch of input data from data set\n",
        "print(test_input)\n",
        "out = model.predict_on_batch(test_input)   #Probability predictions for giving input\n",
        "print(out)"
      ],
      "execution_count": 7,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "[[[[  3477.64]\n",
            "   [ 78900.06]]\n",
            "\n",
            "  [[  3727.94]\n",
            "   [ 75148.44]]\n",
            "\n",
            "  [[  3396.8 ]\n",
            "   [ 72802.24]]\n",
            "\n",
            "  [[  3273.75]\n",
            "   [ 71745.32]]\n",
            "\n",
            "  [[  2800.58]\n",
            "   [ 72721.22]]\n",
            "\n",
            "  [[  2346.93]\n",
            "   [ 75755.16]]\n",
            "\n",
            "  [[  2030.39]\n",
            "   [ 79904.28]]\n",
            "\n",
            "  [[  1509.42]\n",
            "   [ 84915.09]]\n",
            "\n",
            "  [[   743.17]\n",
            "   [ 89481.  ]]\n",
            "\n",
            "  [[   657.4 ]\n",
            "   [ 94311.15]]\n",
            "\n",
            "  [[   935.75]\n",
            "   [ 98981.73]]\n",
            "\n",
            "  [[  1121.12]\n",
            "   [102664.29]]\n",
            "\n",
            "  [[  1213.22]\n",
            "   [105564.29]]\n",
            "\n",
            "  [[  1431.9 ]\n",
            "   [107830.95]]\n",
            "\n",
            "  [[  1514.87]\n",
            "   [109213.87]]\n",
            "\n",
            "  [[  1719.13]\n",
            "   [109550.17]]\n",
            "\n",
            "  [[  1833.03]\n",
            "   [109127.92]]\n",
            "\n",
            "  [[  1804.25]\n",
            "   [107870.04]]\n",
            "\n",
            "  [[  1379.8 ]\n",
            "   [105518.03]]\n",
            "\n",
            "  [[  1177.16]\n",
            "   [101817.23]]\n",
            "\n",
            "  [[  1235.9 ]\n",
            "   [ 99768.25]]\n",
            "\n",
            "  [[  1482.94]\n",
            "   [ 95794.2 ]]\n",
            "\n",
            "  [[  1845.18]\n",
            "   [ 89056.8 ]]\n",
            "\n",
            "  [[  2105.73]\n",
            "   [ 82802.53]]]]\n",
            "tf.Tensor([[1838.3258     750.85504     -5.9897385]], shape=(1, 3), dtype=float32)\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "0GKZXbJxNP9w",
        "colab_type": "code",
        "outputId": "14881a82-10fa-4156-9ef3-15dfdaeb7bc5",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 187
        }
      },
      "source": [
        "# Checking accuracy for the givin input of previous code cell\n",
        "\n",
        "soma = test_input.sum(axis=0).sum(axis=0)\n",
        "media = np.mean(test_input, axis = (0,1))\n",
        "desv_pad = np.std(test_input, axis = (0,1))\n",
        "print(\"Potência gerada pelo vento = \",media[0])\n",
        "print(\"Desvio padrado da potencia do vento = \",desv_pad[0])\n",
        "wls_true = (soma[0]/24)/(soma[1]/24)\n",
        "print(\"Wind Load Share = \", wls_true)\n",
        "pot_vento_pred = out[-1][0]\n",
        "desv_pad_pred = out[-1][1]\n",
        "wls_pred = out[-1][2]\n",
        "print(\"Valor predito da potencia gerada pelo vento = \", pot_vento_pred)\n",
        "print(\"Valor predito do desvio padrao da potencia do vento = \", desv_pad_pred)\n",
        "print(\"Valor predito do Wind Load Share = \", wls_pred)\n",
        "c = 0\n",
        "if (media[0] <= 4000 and pot_vento_pred <= 4000) or ((media[0] > 4000 and media[0] <= 8500) and (pot_vento_pred > 4000 and pot_vento_pred <= 8500)) or (media[0] > 8500 and pot_vento_pred > 8500):\n",
        "  print(\"Acertou a potencia do vento!\")\n",
        "  c = c+1\n",
        "else:\n",
        "  print(\"Errou a potencia do vento!\")\n",
        "if (desv_pad[0] <= 1150 and desv_pad_pred <= 1150) or ((desv_pad[0] > 1150 and desv_pad[0] <= 2300) and (desv_pad_pred > 1150 and desv_pad_pred <= 2300)) or (desv_pad[0] > 2300 and desv_pad_pred > 2300):\n",
        "  print(\"Acertou o desvio padrão da potencia do vento!\")\n",
        "  c = c+1\n",
        "else:\n",
        "  print(\"Errou o desvio padrão da potencia do vento!\")\n",
        "if (wls_true <= 0.075 and wls_pred <= 0.075) or (wls_true > 0.075 and wls_pred > 0.075):\n",
        "  print(\"Acertou o Wind Load Share!\")\n",
        "  c = c+1\n",
        "else:\n",
        "  print(\"Errou o Wind Load Share!\")\n",
        "print(\"Percentual de acerto = \", (c/Yclasses)*100)  "
      ],
      "execution_count": 8,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Potência gerada pelo vento =  [1865.16666667]\n",
            "Desvio padrado da potencia do vento =  [862.8141685]\n",
            "Wind Load Share =  [0.02015267]\n",
            "Valor predito da potencia gerada pelo vento =  tf.Tensor(1838.3258, shape=(), dtype=float32)\n",
            "Valor predito do desvio padrao da potencia do vento =  tf.Tensor(750.85504, shape=(), dtype=float32)\n",
            "Valor predito do Wind Load Share =  tf.Tensor(-5.9897385, shape=(), dtype=float32)\n",
            "Acertou a potencia do vento!\n",
            "Acertou o desvio padrão da potencia do vento!\n",
            "Acertou o Wind Load Share!\n",
            "Percentual de acerto =  100.0\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "0o6bFu0YiP49",
        "colab_type": "code",
        "outputId": "5cb071b2-60cc-4d6c-d118-9347301dc6f0",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1000
        }
      },
      "source": [
        "# Ploting\n",
        "\n",
        "import matplotlib.pyplot as plt\n",
        "\n",
        "plt.plot(history.history['nossa_metrica'])\n",
        "plt.plot(history.history['val_nossa_metrica'])\n",
        "plt.title('Model custom accuracy')\n",
        "plt.ylabel('Custom Accuracy')\n",
        "plt.xlabel('Epoch')\n",
        "plt.legend(['Train', 'Test'], loc='lower right')\n",
        "plt.show()\n",
        "\n",
        "plt.plot(history.history['potvento'])\n",
        "plt.plot(history.history['desvpad'])\n",
        "plt.plot(history.history['wls'])\n",
        "plt.title('Model custom accuracy')\n",
        "plt.ylabel('Custom Accuracy')\n",
        "plt.xlabel('Epoch')\n",
        "plt.legend(['WP Train','StDev Train','WLS Train'], loc='lower right')\n",
        "plt.show()\n",
        "\n",
        "plt.plot(history.history['val_potvento'])\n",
        "plt.plot(history.history['val_desvpad'])\n",
        "plt.plot(history.history['val_wls'])\n",
        "plt.title('Model custom accuracy')\n",
        "plt.ylabel('Custom Accuracy')\n",
        "plt.xlabel('Epoch')\n",
        "plt.legend(['WP Val','StDev Val','WLS Val'], loc='lower right')\n",
        "plt.show()\n",
        "\n",
        "plt.plot(history.history['loss'])\n",
        "plt.plot(history.history['val_loss'])\n",
        "plt.title('Model loss')\n",
        "plt.ylabel('Loss')\n",
        "plt.xlabel('Epoch')\n",
        "plt.legend(['Train', 'Test'], loc='upper left')\n",
        "plt.show()"
      ],
      "execution_count": 9,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYIAAAEWCAYAAABrDZDcAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz\nAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4xLjIsIGh0\ndHA6Ly9tYXRwbG90bGliLm9yZy8li6FKAAAgAElEQVR4nOzdd3iUVfbA8e9JQhJCCqRQQgu9Sgdp\nKggqWHDFigU76q7r6qq7ruvPtkXdta7ouiyiYEPFhi5IsyEgvfdACEmAEBLSe3J/f9wJGULKJGQy\nKefzPHmSeeedNycJvGduO1eMMSillGq6vDwdgFJKKc/SRKCUUk2cJgKllGriNBEopVQTp4lAKaWa\nOE0ESinVxGkiUPWCiESJiBERHxfOvU1Efq6LuJRqCjQRqGoTkUMiki8i4WWOb3bczKM8E5l7iMi7\nIvJXT8ehlLtoIlA1FQNMK3kgIucAAZ4LR9UGV1pkqvHRRKBq6j1gutPjW4F5zieISIiIzBORJBGJ\nFZEnRMTL8Zy3iLwoIidE5CBwWTmvfVtEjopIgoj8VUS8XQlMRMaKyGoRSRWROBG5zXH8BxG5y+m8\nU11MYr0iIsdFJF1EtotIfxGZAdwE/EFEMkXka8f5fRzXSxWRnSIyxem674rImyKy2PGaVSLSVkRe\nFZGTIrJHRAZXEv9rjrjTRWSjiJzn9Jy3iDwuIgdEJMPxfEfHc/1EZJmIpIhIoog87hTPX52uMU5E\n4p0eHxKRP4rINiBLRHxE5DGn77FLRK4qE+PdIrLb6fkhIvKoiHxW5rx/ichrrvzdlOdoIlA19QsQ\n7LghegM3AO+XOed1IAToClyATRy3O567G7gcGAwMA64p89p3gUKgu+Oci4G7qIKIdAYWO753BDAI\n2OLCz3MxcD7Q0xHzdUCyMWYW8AHwD2NMoDHmChFpBnwNLAVaA78FPhCRXk7Xuw54AggH8oA1wCbH\n4wXAy5XEst4RdyjwIfCpiPg7nvs9tiV2KRAM3AFki0gQsBz4FojE/t5WuPBzl5iGTcYtjTGFwAHg\nPMfv4hngfRFpByAi1wJPY/+ewcAUIBn7958kIi0d5/lg/12c9gZB1T+aCNTZKGkVXATsBhJKnnBK\nDn8yxmQYYw4BLwG3OE65DnjVGBNnjEkBnnN6bRvsje5BY0yWMeY48IrjelW5EVhujPnIGFNgjEk2\nxriSCAqAIKA3IMaY3caYoxWcOxIIBJ43xuQbY74DvsGpqwz4whiz0RiTC3wB5Bpj5hljioCPscmt\nXMaY9x1xFxpjXgL8gJIkcxfwhDFmr7G2GmOSsUn1mDHmJWNMruN3vtaFn7vEvxx/ixxHDJ8aY44Y\nY4qNMR8D+4ERTjH8wxiz3hFDtDEm1vH7+gm41nHeJOCEMWZjNeJQHqCJQJ2N97A33ts4811fONAM\niHU6Fgu0d3wdCcSVea5EZ8drjzq6XlKB/2DffVelI/bdbLU4buYzgTeA4yIyS0SCKzg9EogzxhSX\nib+90+NEp69zynkcWFEsIvKIo9slzfGzh2B/n1Dxz1ejn9uJ898CEZkuIlucfv/9XYgBYC5ws+Pr\nm7H/RlQ9p4lA1ZgxJhY7aHwp8HmZp09g32V3djrWidJWw1HsDcX5uRJx2O6UcGNMS8dHsDGmnwth\nxQHdKngui9MHtNs6P2mM+ZcxZijQF9tF9GjJU2WucwToWDLe4RR/AmfJMR7wB2yLqZUxpiWQBojj\nlIp+vjhsF1x5Kv25HU79jI7utf8C9wNhjhh2uBADwJfAABHpj22lfFDBeaoe0USgztadwIXGmCzn\ng44ukE+Av4lIkOPm8ntKxxE+AR4QkQ4i0gp4zOm1R7H97y+JSLCIeIlINxG5wIV4PgAmish1jkHP\nMBEZ5HhuCzBVRAJEpLsjdgBEZLiInOvo/88CcoGSd/yJnH6TXQtkYweQm4nIOOAKYL4L8VUlCDs2\nkgT4iMiT2H74ErOBv4hID8cA9wARCcN2TbUTkQdFxM/xOz/X6ee+VERCRaQt8GAVMbTAJoYkABG5\nHdsicI7hEREZ6oihu+Pvi6MrbAF2bGOdMeZwzX8Vqq5oIlBnxRhzwBizoYKnf4u9qR4EfsbeHOY4\nnvsvsATYih1ELduimA74AruAk9ibSzsX4jmMbaE8DKRgb4IDHU+/AuRjb+xzOf3darAjppPYbp5k\n4J+O594G+jq6Sb40xuRjb/yTsS2fN4Hpxpg9VcXngiXYAd99jjhyOb3b5mVsEl0KpDtia26MycCO\n1VwBHMP26Y93vOY97O/5kON1H1cWgDFmF3Y8Zw32d3UOsMrp+U+Bv2H/nhnYVkCo0yXmOl6j3UIN\nhOjGNEqp2iQinYA9QFtjTLqn41FV0xaBUqrWOMZNfg/M1yTQcOgqQqVUrRCRFtiupFjs1FHVQGjX\nkFJKNXHaNaSUUk1cg+saCg8PN1FRUZ4OQymlGpSNGzeeMMZElPdcg0sEUVFRbNhQ0WxFpZRS5RGR\n2Iqe064hpZRq4jQRKKVUE6eJQCmlmjhNBEop1cRpIlBKqSZOE4FSSjVxmgiUUqqJ00SglIcYY9gY\nm8L8dYfRUi/KkxrcgjKlGrq8wiIWbz/GnFUxbItPAyA5K5/fjO9+xrnFxYafo0/ww94kuka0YGTX\nULpF2F0uM/IKycgtpH3L5nUav2p8NBEoVQfyCotYHZ3M19uOsGxnIhl5hXSNaMFfruzH2pgUXly6\nl36RwYzrZbdlTs3O593Vh/h0QzwJqTn4eAmFxbbVEOTvQ0FRMbkFdgO120ZH8fQUV3bxVKp8mgiU\nqqHs/ELWHEhm5f4TdA4LYPqoKLy95NTze49l8L9tR1gbk8LmuFTyC4sJ8vfhkv5tuXJQJGO6hePl\nJVw9tAPRxzP53fwtfPHr0aw6kMzLS/eSmlPA2O7h/OnS3lzUtw3H0nJZG5PC1rhUAny9aR3kz+5j\n6by7+hD9IoO5dljHSqJVqmINrgz1sGHDjNYaUp606fBJ3vrhAD/sSyK/sBhfHy/yC4sZ0qklL147\nkJYBvry8bC8frj2MiNAvMpgRUaGM7h7GmO7h+Pl4n3HNw8nZXDHzZ7LyCiksNpzbJZSnruhH38jg\nciIoVVhUzPQ569gQe5IF945iQIeW7vqxVQMnIhuNMcPKfU4TgVJVM8aw+kAyb/4QzaroZFoFNGPq\nkA5c2Ls1w6JasXj7MZ78agf5RcX4enuRlV/ELSM787sJPWjVwtel77Eq+gQvL9vHnWO7MLl/W0Sk\n6hcByZl5TJm5CmMMn/96DG1D/M/mR1WNlCYCpWooMT2XBRvj+WRDHLHJ2UQE+THjvK7ceG4nWvj5\nnHHuk1/tIL+wmD9d2oeebYLqLM4dCWlc/e/VGODKgZHcNiaKlgG+7ExIY9fRdOJScjiekUtSRh6+\nPl50DmtBl/AWTOjdmoEdtRXRFGgiUKoSadkFLNgUz5WDIgkP9Dt1fP2hFKa/vY6cgiJGdg3l+uEd\nmdy/Hf7NzuzaqQ8OJmUyZ1UMn21MIKeg6NRxL4G2wf5EBPsTEehHXmERh5KziD+Zg6+3Fx/NGMmQ\nTq08GLmqC5oIlKrA93uP89hn20hMz6NjaHPeuW0E3VsHsiMhjWmzfiEiyI/Ztw6jq2PKZkOQll3A\nl1sS8BLoGxlCn3ZBBPieOS8kOTOPqf9eTWZuIV/8egydwgI8EK2qK5oIlCojMT2Xl5fu4+MNcfRs\nE8i9F3Tj74t2U1Bk+PNlfXh+8R6aN/Pm03tHEdmI5+kfTMrkqjdXExboy7w7RrDh0Em+2XaE7Pwi\n/nxZH/pFhng6RFVLNBEo5bAlLpU5P8ewaPtRio3hngu68eDEHvj5eBOXks1t76zjQFIW4YF+fHrv\nKLqEt/B0yG639mAyt7y9jvwiuy6hXYg/BUWG1Ox8HpjQg/vGdaOZtxYhaOg0EahG7+Wle/lu73Fe\nnzak3Jt3Vl4hTy/cyacb4wny8+G64R25dVTUGd0hadkFvPFDNFcP6UCvtnU32Otpy3clsjYmmUv6\ntWVIp1ak5RTw1MKdLNx6hAEdQnj1+kENqntMnUkTgWrU3vsllv/7cgc+XkKQvw+zbx3G0M6hp57f\nkZDGAx9tJiY5i1+P68Z947oT6KdrKV2xaPtRHv9iO/mFxTx7ZX+uHtLe5Wmtqn7RRKAare/3HOfO\nuesZ16s1f76sD3fN3UBCag4PTexJak4+u46k88vBZMID/Xjl+kGM7Brm6ZAbnKNpOTw4fwtrY1K4\nfEA7/jipNx1DdWC5odFEoBqlXUfSufat1USFt+CTe0bRws+HlKx8ZszbwIbYk/h6e9GzbSDDOodW\na2GXOlNRseHfP0TzrxXRFBnDVYPb8+tx3bS7qAHRRKAanez8Qi7718/k5Bfx1f1jaBNcupq2sKiY\n2JRsOrYKwNdHBzlr07G0XGb9dJAP18VSVGz4/L4xnNNBZxY1BJUlAv1fohqk5xfvIeZEFi9fP/C0\nJADg4+1Ft4hATQJu0DbEnyev6MtPfxhPkH8znlu8W/dSaAT0f4pqcH7al8S8NbHcObYLo7uFezqc\nJql1kD+/vbA7qw8k8+O+JE+Ho86SJgLVoKRlF/Dogq10bx3Io5f08nQ4TdpN53amU2gAzy/eQ1Fx\naasgK6+QlKx8UrLySc8t8GCEylU6h041CPEns1m0/SgLNsaTnJnP7OnD623Nn6bC18eLRy/pxW8/\n2swXmxMY3S2M15bv59ONcTjlBR65uCf3X9jDc4GqKmkiUPXW0bQcFm0/xjfbjrD5cCqAXdx0wyAd\noKwnLjunHf9deZC/fLPLFrozcPPIzqe201wVfYIXl+6jfavmXDW4g4ejhejjGXy15QjTR0UREeR3\nxvPGGKKPZ7I5LpVz2ofQp13pfhDH03N5/btouoS34PYxUY1qPYVbE4GITAJeA7yB2caY58s83wmY\nC7R0nPOYMWaRO2NS9Y8xhhW7jzP754MkpucBdrri4ZRsAPq2C+bRS3px+YB2dA5r/CUfGhIvL+GJ\ny/oyfc5apgyM5MGJPejQqnSNwbQRnZg+Zy1/XLCdyJDmnOtYx5GTX0Rz37pr0R1JzeHV5ftYsDGe\nYgPf7TnO/BkjCfJvBkBKVj5/+WYXP+5LIiUr/9TrJvZpw70XdGX9oZPM/G4/Wfm2qmtCag5/vrQP\nXo4d6Y6n55KYnkfn8ACCHddsSNw2fVREvIF9wEVAPLAemGaM2eV0zixgszHm3yLSF1hkjImq7Lo6\nfbTxKC42fLvzGK9/F83uo+l0aNX8tHLIPVoHctmAdjpXvQEwxlT4Djktu4Cp/17Ficx8RnYNZUdC\nOgmpOVzQM4KZNw4+dTOuDcmZeRxLz+V4Rh4JJ3PYeSSdXUfsngyCcPPIzgzsGMLDn2zl3K6hzLlt\nOAeTsrh73gaOZ+Rx+YB2jOwSxjkdQli6M5E5q2JIy7HjHBP7tOHxS3szb00s764+xNTB7blpZCfm\nro5l0fajp/aUDmvhy6huYTx7ZX9C69HaFY+sIxCRUcDTxphLHI//BGCMec7pnP8AB40xLzjOf8kY\nM7qy62oiaPiMMazcf4J/LNnDjoR0uka04DfjujNlUKQWN2uk4lKyuePd9RQWG/pFBtM6yJ95aw7R\nvXUg79w+nHYhrld4zS8sJju/kJYBp99kX1yyl5nfR592LNjfh36RIQzoGMItIzufaq18tjGehz/d\nysiuoWyLTyPI34dZtww7Y5OezLxCFm45QqfQAMb2sDPUjDHM/C6al5btAzhVu2pY51bEpmRzMCmT\nLzcfISLIjzdvGnLaNStLmO7mqURwDTDJGHOX4/EtwLnGmPudzmkHLAVaAS2AicaYjeVcawYwA6BT\np05DY2Nj3RKzcq/Dydl8v/c4/9t2lHWHUmjfsjkPX9yTKwe1P23Td9U0rNyfxH3vb6KFnze/GtSe\nQ8lZxCZn08LPhxFdQu1HVOhpO8HFnMjivvc3EpeSzSvXD+Lifm0BeGdVDM98vYsrB0UyqV9bWgf7\n0TakOZEh/hXeeN/68QDPL97DwI4tmXXL0DPWo1Tl661HSM3O56ohHc6oXbUtPpX73t9EUkYe1w3v\nwNHUXHYeSafIGL793XmEBZ45PuFu9TkR/N4Rw0uOFsHbQH9jTHFF19UWQcNzODmbGe9tYM+xDAC6\nhLfglpGduWlkp3I3cldNx55j6dw1dwPH0/PoFBZAVFgAKVn5bItPo7DYEOjnw7XDOnDb6Ch2H83g\n0U+34u0tRIY0Z9fRdB65uCedwlrwu/mbubhvG968aajLbyqMMWyJS6VPu2C3zEA7mZXPI59u5fu9\nx+kWEUjvdsEs2n6UG0d04i+/6l+ta8UmZ/HkVzt5+OKeDOhQs61FK0sE7hwsTgA6Oj3u4Djm7E5g\nEoAxZo2I+APhwHE3xqXq0InMPKbPWUtqTgFPXt6X8b1bN4ka/8o1vdsG89Oj4zFw2g08O7+QTbGp\nLNgYx3uOPnljYGCHEN64aQjhgX489tk2Xlxqu2dGRIXy2g2Dq9WyFBEGu3GLzlYtfHn7tuEUFhXj\n4+jyDGnuw4frDnPr6M50b31mmfNNh0/y5eYEercNZnzvCMID/Zj100H+tWI/zby9iD+ZU+NEUBl3\ntgh8sIPFE7AJYD1wozFmp9M5i4GPjTHvikgfYAXQ3lQSlLYIGo7MvEKmzfqF/ccz+PBu3RdX1Uxi\nei4frj2Mt5dwzwVdT7UijTHMXhnDmoPJvHLdIEIC6v9snROZeYz75w+M7BrK7FuHnzq+LzGDF5fs\nZemuRJp5CwVF9hbYMqAZqdkFTO7flqen9Kt295UzjxWdE5FLgVexU0PnGGP+JiLPAhuMMQsdM4X+\nCwQCBviDMWZpZdfURFD/GWM4kJTFM1/vZPWBZGbdMpQJfdp4Oiyl6oV//3CAF77dw4d3n0tYCz/e\n+D6ab7YdoYWvD/dc0JXbx3QhITWH7/ccZ1t8GlcNbs/Evmf//0erj6o6kZSRxxvfR7NiTyJxKTmI\nwAtTB3Dd8I5Vv7i+KC6Gn/4BQ6ZDcKSno1GNUG5BERNe+pGM3ALScwtp4evNzaM6c+/53dxaKt1T\nYwSqCdmRkMbd8zaQnJXP+T3Cuef8bozv3Zr2DW3j9xP74IfnwMsbzn/U09GoRsi/mTdPXdGXpxfu\n5PYxXbh9TNQZU2HrmiYCdda+2XaERz7dSmiAL5/fN5r+7Rtw+Ye0ePv5yBbPxlETmUnQvCV41/++\n8qbu4n5tT019rQ909Y46K19uTuD+DzfTLzKEr+4f27CTAEBanP18ZLNn46iuokKYOQxW/8vTkagG\nSBOBqrGE1Bz+78sdDI9qxYd3n1tuEa8GpyQRpCdARqJnY6mO9HjITYXY1Z6ORDVAmghUjRQXGx75\nZCvFxvDydYMaz8Kwkq4hgKMNqHvo5CH7+chmaGATQJTnaSJQNfLO6kOsOZjMk1f0pWNoQNUvaCjS\n4qHtABCvhtU9dNJRdiU7ubRVo5SLNBGoatsen8YL3+5hYp/WXDesAU0NdUVaHET0hvBeDSsRpDrV\n32pIcat6QROBcpkxhnlrDnH1W6tpFdCMv089p1FtzkFxEaQfgZYdIXJww+pmOXkIgiLBq5kmAlVt\nOn1UueRkVj6PLtjG8t2JjO8VwT+vHUi4ByooulXGMSguhJAOENgWtn4IGUcbxsKyk7EQ3gMCIzQR\nqGrTRKCqtDUulV9/YEvqPnl530a3Td8pJQPFIR3B31HY68jmhpEIUmOh12SQLrDzC9uSaYx/I+UW\n2jWkKmSM4f1fYrn2rTUALLhvFHeM7dI4kwCUDrKGdIC2/UG8IWGTZ2NyRV4mZCVBqyjbpZWbBidj\nPB2VakC0RaBOU1xs2Hkkne/3HmfF7kS2xqcxrlcEr14/yOPL4N3OORE0aw6t+zaMbpbUw/Zzy862\newhs3KFdPReTalA0EahTsvIKuXveBlYfSEYEBnVsyVNX9OXWUVGnNulu1NLibZeQn6NOfOQg2PO/\n+tfNEr8R2g0Eb8d/35IZQ62iIKIPePvZRND/ao+FWC0pB+3vPSDU05E0WZoIFADpuQXcNmcdW+PT\nePLyvlw5KNIj2+l5VFq8HR8oETkYNr9n33G36uy5uJzF/ARzr4ApM2HILfZYyWKyVlHg42u7tRpS\nraR5v4Ko8+BXb3g6kurZ8iF4+cCA6zwdyVnTRKBIzc5n+px17D6azhs3DmZS/3aeDsn98jJg9zdw\nzrVO76zjoGWn0nMiB9vPRzadmQj2LbFdLyVdMeWJ32DPc75e70vPLu6fX7WfD610SgSx0KwFBISV\nfp+tH9uS2l71fBgwP9u2aBIb4KZFP70IPv6aCFTDl5ZTwE2z17I/MZO3bm4iG8gUFcKCO2D/Ulup\n85xr7PG0eOg8uvS8Nv2gRQSs+At0uaC062L7AvjsTmg/FO7+rvzvsX8ZzL8JivIAwe67JHDnUug4\nomZxH9sOB1bYrh/nmkKpsTZRlXRfRQ6G9bMh5UDliao+KGnNJEfXvy64yhTk2gF5Lx/778m7Yd9K\n6/nbBeVO2fmF3PnuevYlZvCf6U0kCQAsfcImAS8fiF5hj+WmQV6aXUxWwscPrnvPDiJ/Mh0K8yFu\nHXz5a2jeChI22v76svZ+C/NvhIhe8IcYeDoV/hQPwe1h4QP2OlUxBgrzTj+26jXwDYTzH7ExlQwS\nnzxku4VKnGrJNICB7pSD9nN+pl3H0VAkR4MphqL80mTWgGkiaKLyCou4572NbDp8klevH8z4Xq09\nHVLdWD8b1v4bRv4a+l4J0cttF0pagn0+pMPp53ceZfvjD62Ez++2N/jgSJjxI/gGwbpZp5+/dzF8\nfLNtTdy6sLQV4RcEl70ESbth9Wvlx5abBt/9Dd6bCv/sBn9rZ7sfiott98+Oz2HobdDL0b0Uu9om\njJOxdsZQifBe0CwADq8561+X2zlPcz2xz3NxVFfSnvK/bqA0ETQyxhg+2xhPYnpuhecUFRse+ngL\nK/ef4PmpA7hsQBMYEyjMg59fgUV/gB6XwMV/he4TIes4JO5wmjpaTu2kgdfb3cp2fWnfAd74ie2K\nGTQNdn5uN4QBSD4An91lB2tv+dK2Gpz1mgT9roIf/wknos/8Pkseh5Uv2ph6TbYf3/3FJpaf/mG7\nTUbeZ6e1+odA7CpbZK4g6/QxDG8fmyx2fAYFObXz+3OXFEf3CkDyfs/GUh1Je21hQtBEoOqffYmZ\nPPzpVp78ake5zxtjeObrnSzafownLuvTsPYTrowxtlZQefYvgzdHwfKnoeckuOZtuxVltwn2+ehl\np68hKM+4x+Giv8DNn0NET3ts+N02MWx613b3fHanvald/77dKaw8k16AZv7w9QO2b7nE0W2w+QPb\nUrn3Z7jyDXudS56Dfd/C5vfhnOtsfF5e0Gm0bRE4zxhyNmS6bWHs/ub045s/sCuP64uTMdCmvx3s\nPtGQEsEeaNXFvnFI2uvpaM6aJoJGZtH2owAs2ZnItvjUM55/4/to5q2J5Z7zu3LXeW5ccFRcbAdV\nnW927rT0CXiuI3x1v52tk5cBG96BWePgg2vsu+mbPoNpH5auEwhqY0tOR6+wA8VezWyNofJ4ecGY\nB6CD097fET2h63hYP8cmmSOb4cqZFSeTku856QX7bn7J4/aYMbD0z7YF4bxPsgiM+jXc+rVNWuc/\nUvpc59G2nzpunX3cssyspqjzbHLYNLf02Ilom4CWP1PJL7KOpcQ4Zl91b2CJYK+tUhvRq1G0CBr2\nULc6w+IdRxnQIYS4lGxeXLqPeXeUzlD5eP1hXly6j6mD2/PHSb3dG0jsz/YdcrOAs58y6Sw3zSaX\nFmGlx+LWwZo3oO05th9983v2nXlxoe1GufRFGHKrnWNfVveJdntH30Db91/d6ZYjZsD8afDLGzDs\nDuhzRdWvGTTNdketmWln9QS3t+sDLn2x/JZE1Bj74ayz4/HWj+znstNbvbxg8C22ayn5AIR1g2VP\n2t/JyRjHmolKElZdKCqwLbH+U203S/w6z8bjqsJ8OyOr92W2RXjoZ9sa9Wq4mzNpi6ARiT6ewb7E\nTKYObs9947rx074k1sWkAPDemkM89vl2zu8ZwQvXDHD/SuFkRx94yoHau2ZhPsyZDDOHwtGtpccW\nPmBvprcvgof3wOWvwoh74K4VcN9qGHF3+UkAbCIoLrSDxuWND1Sl5yUQ2s2u6L3k766/7qJnbT/+\n4j/ANw/aAd6ht7v++nYDbHfKsW12iqtvizPPGXSTvcFuft8mmr3/K11tHLPS9e/lLmlx9nffqguE\n97TrOOr7mAbYmU7FhaUtgsLc0hlcDZS2CBqRxdvt9LtJ/dsR0rwZ/10Zw4tL9zK6WxivLt/PxD5t\nmHnjYJp510H+T4k5/XNtWP0aHN8JzUPt6tpbvrTdOkm7YdrHpV0+w6pxQ+04AvyCIS+9Zu+Qvbzh\nruV2YVGz5tV73dT/wjuT7PqAK9+o3lx072Y29oPfn9ktVCK4nR0Y3/KBHScJ6WRnQB343s6CGjTN\n9e/nDiX/NkK7gG8AYGzrpW1/j4ZVpZKuoIhetkUAtqsotIvnYjpL2iJoRBbtOMbQzq1oG+JPc19v\n7h/fnXUxKby6fD/XDO3AWzcPwb9ZHTVfS6YF1lYVzBPRdrZN31/BjB/srJl5v7KzafpdZWfk1IR3\nM+h6gf26pl0lAaGOG1k1+QXaZHbDR7ZlUl0li9/KDhQ7GzIdMhMhcTtMfMrGGTXGthA8velOyb+N\n0K4Q5lj41hCmkCbtBcS2YsIdEwca+DiBJoJGIuZEFruPpjO5f+lg5w0jOjKyayi/Gd+Nf14zAJ+6\naAmUSDnk+FwLicAY233i4w+T/2H7w29bBAGt7LvwSS+c3fVLbsItPTCDqkW4HUOpyYraU4mgkjpI\nPS623WYdRpR2C0Wdb7tlPL0QKiXG/k0D20JYd3ssuZxptfVN0h5bisQ3wI7pBLVr8DOHtGuokVi8\nw84WmnxO6ZoAPx9v5s8YVffBGFO6YjT1sB0U9G5Ws+ucjIFtn9iujCtes7NuwN6071lpZwcFneWK\n6F6Xwdb50Hns2V2nrrUfBt0uhO4XVXyOt48ta9EsoDTZdDnPfj600rPdGSUror287E01pGPDmDl0\nYp8dHyjRCGYOaSJoJBZvP02o/C0AACAASURBVMagji1p37Ia/dTukpVkFzm1Pcf2f6fFVa82fnER\nfPMQ7PzSln0A6DoOBk8//Tz/YPtxtgIj4I5vz/46da2ZP9ziwpqAsl1eEb3tAHPMStt1VFZRAeSc\nhMAarDbPPA4Lfwu9L7eD1ZXNwko5aAeKS4T3qP9dQ0WFNll1u7D0WERv2PRew6qVVEaVfQUi8pKI\n9KuLYFTNHEjKZHtC2mndQh5V0h1UsmCrut1DS5+w8997TbKtgBk/2DUA9b2SZkMhAlFjbYug7DiB\nMfDJrfByH/j2cTtdtzq++6tdALfwfnh7oq3HVB5jbIvAuUUS1qO0+Fx9lRprCwmWbREUZJVuddoA\nufI/azcwS0TWisi9IhLi7qBU9byzKgZfby+mDvHwvPASJYOAJe+aqjNgvP5t+OVNOPdemDrL1taJ\nHNzgqzvWO1HnQcZRO0vH2frZdppphxH27/D6MFvS2hXHdtg1HOfeB1f9x04H/e8EeP9q2LXQtjRK\nZCZCQfaZLYL8TBtXfXVqxpBzInB83YDHCapMBMaY2caYMcB0IArYJiIfish4dwenqpaanc9nGxO4\nclAkEUH1ZCOZlBhAoOO5tmSyqy2C6BWw6FE7wFmdOfmq+rqcbz8f+qn02LEdsOTPdszhtv/BjO9t\nH/4XM+yMrcoYY1dK+4fAuD/CwBvgtxvhgj9C4i745BZ4ua+dugpOU0edugxLSmbX53GCU4mgZ+mx\nU4mg4Y4TuNTWFhFvoLfj4wSwFfi9iMx3Y2zKBR+tiyOnoIg7xtajOcwnYxz7/vrbG4krs1NSD8OC\n2+1/qmvmNOhVmg1CWHc722XLR7Bnke2vX3CHvZH/6t+2Gy5ysB07GXADfP9X+P65irtt9i2BmB/h\ngsdKi+35B8P4P8FDO2yhPv8QW8I7N91p6miZriGo3+MESXshuEPpmhWw04dbRDToRFBle1tEXgEu\nB74D/m6MKVkH/oKIVNoWEpFJwGuANzDbGPN8OdcuaVkEAK2NMRVU61JlFRQVM3f1IUZ3C6NPu1oY\nNK0tKTGlc9tDu1TdIigqhM/utvWJbnj/9P9kyj1EYOA0W5F1vtPCslu+sIPnJby84Vdv2pIdPz5v\nV9Re+MTpg6KF+XZcJ6w7DL/zzO/l5W1XYLcIt11FK56xiwLF6/TV3MGRdrV04s7a/3lLGAOHf7E1\no1yZyXZ8t60oWzKttTDv9IHiEhG9bdnvglz7Bqis/Gz7cxUXnL75UUWKi2zLKOWgnd5c0cr4WuJK\nx+s24AljTFY5z1W41ZKjFfEGcBEQD6wXkYXGmF0l5xhjHnI6/7fAYFcDV7bA3LH0XP52VT1biZly\n0NZhAdv0L1m8VNGMip/+AXG/wNTZ1ZtdpM7OxKfgvIdt3aMjW2wCKO8m5+UNU163n1e+aP++U163\nC+IyjsGnt9kS0tM+rvzm2n6oHftZ+5YdYA3pcPoNTgQ6jYSN79hB5iHT7TaQ/rU4LLnuv7D4URh0\nsy0QWPJvMjvFdkuG94DBN9vYdnxuixj6trA1pUrO7XPlmdc99174+Cb46jdw9Wx7rjGwaR788m84\nsdduZAO29Ilz8cIT0bZbLT/TPi7Isa2Lgmz7ePQDcPFfau93UA5XEkGq83ki0hIYZ4z50hhT2ZSC\nEUC0Meag43XzgSuBXRWcPw14yqWoFcYY5vwcQ5fwFvVrU5m8DMg+Udrkb9XF/oPOTISgcmY1HfoZ\nfvonDLwRBlxbt7EqezPvNNJ+VMbLy87gCu0CK56175TPf8SOKeSlw9Vvu7a6+8InYM839kbX5YIz\nn7/2Xdj2MWycC4segaX/Z1eOD5luYyz7ZuLwL5CwyRb7q2pBYFq8bY0EhMOW9+1Nf+yDkHUC5l1p\nYyouhB9fsN1iCRvtONe1c225jsr0uRwmPGl/N+E9YMyDsOhhW+epw3BbVbZNf/szLXkc7lhif5bi\nYvjyPvu92w201/IPtkUS2w20W5OuecNup1ryvBu4kgieMsacmqxsjEkVkaeAL6t4XXsgzulxPHBu\neSeKSGegC7b7qbznZwAzADp16lTeKU3O3sQMtsan8cyUfu4vIFcdJd1AJbNBShJCSsyZiSAvEz6/\nx5576T/qLkZVMyIw9iFoN8iOJ3x2p/3b3fK53ZHNFX6Bdqe2D68rfzGbf7AtEjj8LlvWe9NcW858\n64e2MN+Q6bZLqyjfVlPd/ol93ZLHofsEewPtNfnMlokx9h1/cRHctcxOc13+tG1trP2PHbO4aYGN\nafP7dpbTuffaPShc7ZYZ+3v77v6H52wySzkI5/8Bxj1WOuaVm2rXWez8wlZd3TjHVl391Vvl137q\nNQkO/mgLK961wm2z51y5ankDyrUdzQ3AAmNMuTuLGGNmAbMAhg0bVo8nGdedzYftXgPjekVUcWYd\nKzsIWJIQTsbYbR+drXoN0uPhjqU6LtCQdBsP9/wI2z+FYXdWvAlPRXpeYovsdRhe8Tki0H6I/bj4\nb/bGuWmu3bdhxTOlZcbPe8R2H21fYG/gn9xiB24H3Wi7f8J72GvtXgh7F9kbe2hX+/1TD9vSJc0C\n7GB2Sc2pC5+wH9UlAle8aq97bBtMm2+TkrNBN8HaWbD8KYgcBMuetoslB95Q/jWbt4LJL9iJFGvf\ngtH3Vz8uF7hyQ98gIi9j+/sBfgNUsErkNAmAc1utg+NYeW5wXFe5aFt8GsH+PnQKrUGxM3cq2yJo\n2ckOCpYdME6Lh9WvQ7+p0KnchqKqz1p2suMLNTX4ZtfP9QuEIbfYj8Rdtt89L912TZWMKV34Z/vO\nO3q57VZaPdO+0WgeartUEnfYTYhG/tqe36w53PAhLP6jbYG4MoDrCh8/mP6l7ecvb9W7lzdc8lfb\nFTXbUQL98lcqX5Hc7yrbwvj+b7YLqrIigzUN24Vzfgv8H1CyqmQZrt201wM9RKQLNgHcANxY9iQR\n6Q20AhrATtv1x/aEVAZ0aInUtyXtJ2MgIKz0P4GPr51uV3ZR2Ypn7eDZxKfrOkLVkLXpC5OfL/+5\nktlJPS+B9KN2YdyRLXbvCmPsALdz10pga7j2ndqP0btZ5YPmXcdBz8mwbzFMfKbqCRIitjvtzVF2\nrU15M7POUpWJwDFb6LHqXtgYUygi9wNLsNNH5xhjdorIs8AGY8xCx6k3APONqc/ryuuX3IIi9h7L\ncO9WkzVVsvWgs7JTSBM22nc4Yx+qvHKmUjUV3M6OM9RXl78CO8+3rRFXhHSAB7acvjNfLXJlHUEE\n8AegH3Bqgqwxppx5ZqczxiwCFpU59mSZx0+7GKty2Hssg4Iiw4D29bDaR0rMmTNQQrvA7q/t10UF\ntoZNiwg7uKZUUxTczu5HXR1uSgLg2sriD4A92Fk9zwCHsN0+ykO2J9hZu+d0qGeJoDDfDv6WnQ3S\nqgtkJ9sFMnOn2DUDE5+pncqhSqmz5soYQZgx5m0R+Z0x5kfgRxHRROBB2+PTCG3hWz9KTjtLPWz7\n/VuVSQQliWHWOPv81W/bedFKqXrBlURQUjLwqIhcBhwBQt0XkqrKtoQ0+rcPqR8DxTmpdlpbwkY7\n7xsgrNvp55TUkGkRATd84Pqcc6VUnXAlEfzVUXr6YeB1IBh4qPKXKHfJLShiX2IGE3rXk9XE3zxk\n53i37mNronQcYXfOctamL9z8mT1e3TnnSim3qzQROOoF9TDGfAOkUVogTnnIrqPpFBWb+jE+sG8J\n7Pwcxj1uSw9Xpiabsyul6kSlg8WOlb7lrHtWnrI93g4UD/B0IsjLhP89bKsujtUGolINmStdQ6tE\nZCZ2QdmpCqTGmE1ui0pVaFt8GuGBfrQNLqfUbXVknbCLXmpa2fH7v9nVwXcscXuJXKWUe7mSCAY5\nPj/rdMwAVa4jULXPrig+y4HilBiYPQG8msF186pf4iHmJztAPPxOLQ+hVCPgyspiHReoJ7LyCok+\nnsnk/lWUxK1Mbhp8eL2tceIXBO9eBpOes6swyyaXpL02aXQbb2uoGGP3tP32T3b18IQny/8eSqkG\nxZWVxeX+bzfGPFveceU+2+LTKDZwTnVWFGccA/G2m44UFdpNRFIO2J2o2p4Dn8+wNdL3L7Obb3Qb\nb/cU+OE5u4mHKbKFuwZOs/sMbPsYelxiN5avzQ1DlFIe40rXkPPOZP7YbSt3uyccVZnFO47i5+PF\nyG4uLjXPz4KZIyAvDYLb22Jwx7bZ4lslm5dP+xhWvWo3v/jgart1YEEO5KTA0Nuhx0Ww9SNYN8u2\nIsY9bjfZ8HJpu2ulVAPgStfQS86PReRFbCE5VYcKior537ajTOzbhkA/F7eD2PetTQLn3mtLPCTu\nst05Q6aXnuPlBef9Hkbdb+u1b37PHp/wZOmOSL0mQ2aSLf1bdrGYUqrBq8kGMwHYvQVUHVoVfYLk\nrHymDIx0/UU7PofANnDJ30t3SKqIjy/0+5X9KE9gxOmbmiulGg1Xxgi2Y2cJgS0nHcHpM4hUHVi4\n5QhB/j6u70iWl2H7/YfeVnUSUEo1aa60CC53+roQSDTGFLopHlWO3IIiluw8xuUDIvHzcfGmvncx\nFOXZ3Y2UUqoSroz4tQNSjDGxxpgEoLmI6OTxOrRi93Gy8ou4clA1u4WCIqGj/qmUUpVzJRH8G8h0\nepzlOKbqyFdbEmgd5Me5XSuYLVRcbOf3pxy0j3NS7d6t/a7S2T1KqSq50jUkzttIGmOKRaQmg8yq\nBtKyC/hhbxI3j+yMt1cFq4kPfGfr/viF2Pn9OSlQXAD9p9ZtsEqpBsmVG/pBEXmA0lbAr4GD7gtJ\nOVu66xj5RcVMqaxbaNt8u7irZWf46HoIbAshnaD90LoLVCnVYLnSb3AvMBpIAOKBc4EZ7gxKlVq6\nK5F2If4MrKjaaF4G7P4G+k2FO5fCwBsh8xj0v+rMkhFKKVUOVxaUHQduqINYVBk5+UWs3J/EtUM7\nVlxkbvfXUJgDA2+AZs3hV2/CkFsgcnDdBquUarCqbBGIyFwRaen0uJWIzHFvWArsIrLcgmIu6tum\n4pO2zrddQiWzg0Sg82ibFJRSygWudA0NMMakljwwxpwE9O1mHVi2K5EgPx9GVjRbKC3BloQecL12\nAymlasyVROAlIq1KHohIKDUrTaGqoajYsGJPIuN6t8bXp4I/044FgLHdQkopVUOu3NBfAtaIyKeA\nANcAf3drVIotcSc5kZlfRbfQx3ZDeC0Ep5Q6C64MFs8TkQ2U7kg21Rizy71hqaW7EmnmLRXXFjq+\nB47vhEtfrNvAlFKNjktdPI4b/y4RaQFMFZF/GmMuc29oTduyXYmM7BpGsH+z8k84tNJ+7nFx3QWl\nlGqUXJk15CsiVzm6ho5iWwZvuT2yJuxAUiYHk7Iq7xaKXQXBHaBlp7oLTCnVKFXYIhCRi4FpwMXA\n98A8YLgx5vY6iq3J+mRDHCIwsU8FicAYiF0NXS7Q2UJKqbNWWYvgW6ArMNYYc7Mx5muguG7CarqO\npeXy7qpDXDWoPZEtK1gLkHIQMhOh86i6DU4p1ShVNkYwBLuieLmIHATmYzemUW70r+/2U2wMD13U\ns+KTYlfZz53H1E1QSqlGrcIWgTFmizHmMWNMN+ApYBDQTEQWi4hLtYZEZJKI7BWRaBF5rIJzrhOR\nXSKyU0Q+rNFP0UjEnMji4/Vx3DiiEx1DAyo+MXa13Yg+vJJkoZRSLnKpWL0xZrUx5rfYvYpfAUZW\n9RoR8QbeACYDfYFpItK3zDk9gD8BY4wx/YAHqxd+4/Lysn34+Xhx/4U9Kj8xdrUtI6HjA0qpWlCt\nXUuMMcXGmKXGmDtcOH0EEG2MOWiMycd2LV1Z5py7gTccZStKCtw1SetiUvh66xHuHNuFiCC/ik9M\ni4fUWO0WUkrVGneWimgPxDk9Lilh7awngIiswo4/PG2M+daNMdU7J7PyeWnZXj5ce5g2wX7cfX7X\nyl8Qu8Z+7jza/cEppZoET9cM8gF6AOOw3U4/icg5zkXuABxjEjMAOnVqHPPmM/MK+eCXWN784QCZ\neYVMHxXFgxN7VLyArETsKvALhjb96yZQpVSj51IicBSd6+h8vjFmUxUvS3C8pkQHxzFn8cBaY0wB\nECMi+7CJYb3zScaYWcAsgGHDhhkasIzcAt5ZdYg5q2JIzS7g/J4R/PnSPvRqG+TaBWJXQ6eR4KUT\nuJRStaPKRCAifwFuAw4AJTdhQ2ntoYqsB3qISBdsArgBuLHMOV9iF629IyLh2K6iRr0N5u8/2cqy\nXYlM7NOG+y/szqCOLat+UYnMJDixFwZNc1+ASqkmx5UWwXVAN8eAr8uMMYUicj+wBNv/P8cYs1NE\nngU2GGMWOp67WER2AUXAo8aY5Or9CA1H9PEMlu1K5IELu/P7i3tV/wK6fkAp5QauJIIdQEug2jN6\njDGLgEVljj3p9LUBfu/4aPTe/jkGXx8vpo+OqtkFDn4PvkEQOaRW41JKNW2uJILngM0isgPIKzlo\njJnitqgaoROZeXy2KYGrh7QnPLCS6aGVOfgDdDkPvD09xq+UakxcuaPMBV4AtqO1hmrsvTWx5BcW\nc+fYKqaHViQlBk4egpG/qdW4lFLKlUSQbYz5l9sjqYdy8oto7nv2s3NyC4p475dYJvRuTffWgTW7\nyMEf7Oeu4846HqWUcubKyuKVIvKciIwSkSElH26PzMP2HEtn4DNLWbk/6ayv9fmmBFKy8rnrvBq2\nBsCODwS3h/Aqyk8opVQ1udIiGOz47FxfyJXpow3arJ8Okl9UzIZDJzmvRwXbRbogITWHl5buZWCH\nEEZ2Da3ZRYqLIOYn6HWp1hdSStU6V/YsHl8XgdQniem5fL31CAD7j2fU+Dq5BUXc+95G8gqLefn6\nQUhNb+JHt0LOSeja5P4USqk64MpWlSEi8rKIbHB8vCQiIXURnKe8u/oQRcWGPu2C2ZeYWaNrGGP4\n8xc72J6QxivXD6JbRA3HBsBpfOCCml9DKaUq4MoYwRwgA7uw7DogHXjHnUF5UpajBtAl/dpyYe8I\nDp3IIr+w+pOl3l97mM82xfPAhB6V7z3sioPf29pCga3P7jpKKVUOVxJBN2PMU45y0geNMc9gt7Bs\nlD7dEEd6biF3ndeVnm2CKCw2xJzIqtY1cguKeGnpXsZ2D+fBCWc5uJufDYd/0dlCSim3cSUR5IjI\n2JIHIjIGyHFfSJ5TVGx4e1UMQzq1ZGjnVvRobQvB7Uus3jjBkp3HSM0u4L5x3fDyOsvB3bi1UJSv\n4wNKKbdxZdbQvcA8p3GBk8Ct7gvJc9YcSCYuJYfHJvUBoGtEC7wE9lczEXy49jCdQgMY1TXs7INK\ncdTga9Pv7K+llFLlcCURpBtjBopIMIAxJt1RUbRBWbYrkQUb4ziekUdSRh4+XsJXvxlLSEBp/f+V\n0Uk08xbG97bTRf2beRMV1qJaA8YHkzJZG5PCHyb1OvvWAEBmIiDQouZTWJVSqjKudA19BjYBGGPS\nHccWuC8k90jJyuNgUhYBvt70bRfMoeRsvtubeNo5q6JPMKRTKwJ8S/NjjzaB7KvGFNL56+Pw8RKu\nGdqhdgLPTLRJQOsLKaXcpMK7i4j0BvoBISIy1empYMDf3YHVtuuHd+L64XZ3s+Jiw7nPrWD57uNc\nNdjesFOy8tl5JJ3fT+x52ut6tgli+e7j5BUW4edTebmJvMIiFmyM56K+bWgdVEu/ooxECDzLWUdK\nKVWJyt5m9gIux5agvsLpeAZ20/kGy8tLuLBXaxZtP0p+YTG+Pl6sOZCMMTC6e/hp5/ZoE0RRseFg\nUhZ92gVXet2lOxNJycrnhhG1uJ1mZiIEaSJQSrlPhYnAGPMV8JWIjDLGrKnDmOrEhD6t+XhDHBsO\npTC6ezg/R58g0M+HgR1OXyvXs41dCLYvMaPCRGCMYc2BZF7/bj/tWzbnvDLJ5KxkJkLrPrV3PaWU\nKsOVMYKrRCRYRJqJyAoRSRKRm90emZuN7RGOr48Xy3fb/XZWRZ9gZNcwfLxP/5V0CW+Bt5dUOIX0\nm21HmPzaSm6cvZYTmfn83+V9ameQGKC4GDKPa9eQUsqtXEkEFzsGiS8HDgHdgUfdGVRdCPD1YXS3\nMFbsSSQuJZvDKdmM7X7mdE8/H2+iwgLKnTn0yfo47v9wMwD/uHoAqx+7kEn929VekDknobhAE4FS\nyq1cSQQl8ysvAz41xqS5MZ46NaFPG2KTs5m7+hBgWwnl6dkm6Iy1BCt2J/KnL7ZzXo9wFt4/luuG\nd8S/2dnvXXCazGP2s44RKKXcyJVE8LWI7AGGAitEJALIdW9YdWNCb1u7Z+6aQ7QJ9quwMFyPNkHE\npmSTW1AEwMbYk/zmw030iwzmrZuH4uvjyq+xBjId01u1RaCUciNXylA/JiL/ANKMMUUikgVc6f7Q\n3C+yZXP6tAtm99F0xnQPr7BMdM82gRgDX25OYPPhVBZuPUKbYH/m3DacFn5unN+foYlAKeV+Vd7F\nRGS609fOT81zR0B1bWKf1uw+ms7YSmb69Gxjaw499vl2Any9uWJgO343sWfNN6F3lbYIlFJ1wJW3\ns8OdvvYHJgCbaCSJ4LphHdl9NIMJvSu+2XaPCOT+8d3pGNqcywZEEujOVoCzzETwDQS/s9jLQCml\nquBK19BvnR+LSEtgvtsiqmMdQwOYfeuwSs/x8hIeuaRXHUXkJDNR9yBQSrldTUY5s4AGV3SuQcpI\nhMC2no5CKdXIuTJG8DV2s3qwiaMv8Ik7g1IOmcfszmRKKeVGrnR2v+j0dSEQa4yJd1M8ylnmceiu\nLQKllHtVVn20O9DGGPNjmeNjRMTPGHPA7dE1ZfnZkJeuYwRKKberbIzgVexG9WWlO55T7nRq6qi2\nCJRS7lVZImhjjNle9qDjWJTbIlKWriFQStWRyhJBy0qea17bgagyMrTOkFKqblSWCDaIyBkb0IjI\nXcBG94WkADtQDNo1pJRyu8pmDT0IfCEiN1F64x8G+AJXuXJxEZkEvAZ4A7ONMc+Xef424J9AguPQ\nTGPMbJejb8wyj4F4Q8CZpbGVUqo2VbZDWSIwWkTGAyWT2f9njPnOlQuLiDfwBnAREA+sF5GFxphd\nZU792Bhzf/VDb+RKVhV7uamyqVJKObhSYuJ74PsaXHsEEG2MOQggIvOxVUvLJgJVngwtL6GUqhvu\nfLvZHohzehzvOFbW1SKyTUQWiEjH8i4kIjNEZIOIbEhKSnJHrPVPppaXUErVDU/3O3wNRBljBgDL\ngLnlnWSMmWWMGWaMGRYREVGnAXqMFpxTStURdyaCBMD5HX4HSgeFATDGJBtj8hwPZ2N3QVPFRZCV\nBEHaIlBKuZ87E8F6oIeIdBERX+AGYKHzCSLivNP7FGC3G+NpOLJOgCnWxWRKqTrhth1WjDGFInI/\nsAQ7fXSOMWaniDwLbDDGLAQeEJEp2GJ2KcBt7oqnQSnZtF4TgVKqDrh1qy1jzCJgUZljTzp9/Sfg\nT+6MoUE6tZhME4FSyv08PVisypPmmGyl5SWUUnVAE0F9U1QAv7wFoV0hpNzZtEopVavqaBd25bKN\n78KJvXD9B+Dl7elolFJNgLYI6pOcVPjhOYg6D3pf5ulolFJNhCaC+mTli5CdApf8DUQ8HY1SqonQ\nRFBfpByEtf+BQTdBu4GejkYp1YRoIqgvfnrRlp2+8AlPR6KUamI0EdQHaQmw7RMYMh2C21V9vlJK\n1SJNBPXBL2/akhKjfuPpSJRSTZAmAk/LSbVTRvtPhVadPR2NUqoJ0kTgaRvehvxMGP2ApyNRSjVR\nmgg8qSDXriLuNgHaDfB0NEqpJkoTgSdt/QiyjsOY33k6EqVUE6aJwFMK82HlyxA5BLqc7+lolFJN\nmCYCT9nyPqQdhvF/1lXESimP0kRQm4qLwJiqzyvMswvIOoyA7hPcH5dSSlVCE0FtyU2Df4+BxX+s\n+txN8yA9AcY/rq0BpZTHaSKoSMYxmDMZEndVfa4x8M1DkLQbNr4DmUkVn1uQAytfgk6joeu42opW\nKaVqTBNBRVa+DIdXQ/Tyqs/d8iHs+MwWjCvKh01zKz53wzuQcVRbA0qpekM3pilPWoJ9Zw+QtLfy\nc09Ew6JH7R4CU163XT4b5sCYB8G7zK835yT89A/ocgF0Oc89sSulTikoKCA+Pp7c3FxPh1Jn/P39\n6dChA82aNXP5NZoIyrPyJdvdE94TkvZUfJ4x8MUM8PGFqbPsjmIj7oH502Dv/6Dvlaef/+M/bUmJ\ni//q3viVUgDEx8cTFBREVFQU0gRa4MYYkpOTiY+Pp0uXLi6/TruGyko9bAdzh9xi+/CT9lY8Eyhu\nHSRshIlPQ3CkPdbzEmjZCdb99/Rzkw/Aulkw+GZdRaxUHcnNzSUsLKxJJAEAESEsLKzaLSBNBGX9\n9KLtuz/vYYjoBfkZtk+/PJvmgW8g9L+m9JiXNwy/Cw6thMSdpceXPQnevrrfgFJ1rKkkgRI1+Xk1\nEThLiYEtH8DQ2yGkA4T3ssfL6x7KTYedn0P/q8Ev8PTnBt8CPv6w4E5Y+n/w86uw5xs47yEIauv+\nn0MppapBE4Gzn14ELx8Y+5B9HNHbfi5vwHjHAijIhiG3nvlcQChc9jI084e1b8HypyCkI4y6332x\nK6XqneTkZAYNGsSgQYNo27Yt7du3P/U4Pz/fpWvcfvvt7N1bxaSVs6SDxSWSD9gicOfeW7pLWItw\naB5afotg0zxo3Q/aDyn/eoNvsh+F+fb1LcKhWXP3xa+UqnfCwsLYsmULAE8//TSBgYE88sgjp51j\njMEYg5dX+e/L33nnHbfHqYmgxI8v2D78sQ+WHhOxrYKyLYKj2+DIZpj0QtVrAXx8dXBYqXrgma93\nsutIeq1es29kME9d0a/ar4uOjmbKlCkMHjyYzZs3s2zZMp555hk2bdpETk4O119/PU8++SQAY8eO\nZebMmfTv35/w8HDuvfdeFi9eTEBAAF999RWtW7c+65+j6XQNHfgOvrq//BlASXth+6cw4m4ILPNL\njegFx3ef/rrN74G3yX/KTwAAC2hJREFUHwy4zr0xK6UarT179vDQQw+xa9cu2rdvz/PPP8+GDRvY\nunUry5YtY9euM6sapKWlccEFF7B161ZGjRrFnDlzaiWWptMiSDlob+DjHy+d6lnixxfAp3n5+wJE\n9IbcVMhKskmiIAe2fQx9p9ixAKVUg1CTd+7u1K1bN4YNG3bq8UcffcTbb79NYWEhR44cYdeuXfTt\n2/e01zRv3pzJkycDMHToUFauXFkrsTSdFsGpgd8y/f1J+2DH53DuPbYf/4zXlZk5tPtrW2Bu8C3u\ni1Up1ei1aNHi1Nf79+/ntdde47vvvmPbtm1MmjSp3LUAvr6+p7729vamsLCwVmJpgomgTH9/zI+A\ngWF3uPa6TfOgVRdbUkIppWpBeno6QUFBBAcHc/ToUZYsWVKn37/pdA21CIeAsDNbBMe2Q/NWdt1A\neYLagl+IfV3yAbtQbMKTUMEIv1JKVdeQIUPo27cvvXv3pnPnzowZM6ZOv78YVzZSqenFRSYBrwHe\nwGxjzPMVnHc1sAAYbozZUNk1hw0bZjZsqPSUis2ZDBi449vSY/+dYKd13vZNxa+bfRH4+EH7obD6\ndXhoZ+kUU6VUvbV792769Onj6TDqXHk/t4hsNMYMK+98t72tFRFv4A1gMtAXmCYifcs5Lwj4HbDW\nXbGcUnYGUHERHN8Fbfq78Lpdttx0z0s0CSilGhV39m+MAKKNMQeNMfnAfODKcs77C/AC4P46sc4z\ngABOHrKrg9tUMZsgojdkJ0PWcRgy3e1hKqVUXXJnImgPxDk9jnccO0VEhgAdjTH/q+xCIjJDRDaI\nyIakpEp2/6pK2RlAiTvs57ZVtQgcA8aBbaH7RTX//kopVQ95bMRTRLyAl4GHqzrXGDPLGDPMGDMs\nIiKi5t+07AygYztAvEqPV6S1o69t8E1nbjajlFINnDvvaglAR6fHHRzHSgQB/YEfHGVT2wILRWRK\nVQPGNeY8AwhsmeiwHlXXAAppD7f9zw4WK6VUI+POFsF6oIeIdBERX+AGYGHJk8aYNGNMuDEmyhgT\nBfwCuC8JgKN2UK/SFkHi9qrHB0pEjdWicUqpRslticAYUwjcDywBdgOfGGN2isizIjLFXd+3ShG9\nbIsgN83uRuZqIlBKqWqqjTLUAHPmzOHYsWNui9OtHd7GmEXAojLHnqzg3HHujOWUiN625tChn+3j\ntufUybdVSjU9rpShdsWcOXMYMmQIbdu6Z2OrpjfyWTIwvOMz+1lbBEo1DYsfs5UEalPbc2Byuetk\nqzR37lzeeOMN8vPzGT16NDNnzqS4uJjbb7+dLVu2YIxhxowZtGnThi1btnD99dfTvHlz1q1bd1rN\nodrQBBOBYwrp3sXg3xKC21d+vlJK1bIdO3bwxRdfsHr1anx8fJgxYwbz58+nW7dunDhxgu3bbcJK\nTU2lZcuWvP7668ycOZNBgwa5JZ6mlwhCOtgN5/MzIXJI1RvLKKUahxq+c3eH5cuXs379+lNlqHNy\ncujYsSOXXHIJe/fu5YEHHuCyyy7j/9u7+xC5rjKO498fm7WTtiFJGwk1U92VRiUa+0KR+IJoXEJb\nxQgaailYkg1CEY0iakUhmvhPRHyJLg1pq0YprTGpbSikWLdFBTV9WeM2bapNa9xuTMzuamIUadPm\n8Y97th02u02mO7Oze8/vA8Pee2Zm5zw8wzxzz7lz7ooVK6akP/kVAgkWvAn+3nfmH5KZmTVBRLBm\nzRo2btx42n39/f3s3r2bnp4edu7cydatW5venzyX0BydJ/D8gJm1QFdXF9u3b2d4eBgozi4aGBhg\naGiIiGDVqlVs2LCBvr4+AObMmcOJEyea1p/8jgjg5XmCMy02Z2bWBEuXLmX9+vV0dXVx6tQp2tvb\n2bJlC21tbXR3dxMRSGLTpk0ArF69mrVr1zZtsripy1A3w6SWoR51/BA8fAu8/6teMsKsxLwM9cte\naRnqPD8F5y6Crq+1uhdmZtNCnnMEZmb2EhcCMyu1mTb8PVmvJl4XAjMrrUqlwsjISDbFICIYGRmh\nUqnU9bw85wjMLAvVapXBwUEmdUGrGaZSqVCtVut6jguBmZVWe3s7nZ2dre7GtOehITOzzLkQmJll\nzoXAzCxzM+6XxZKGgL+9yqcvAIYb2J2ZIse4c4wZ8ow7x5ih/rjfEBGvHe+OGVcIJkPSIxP9xLrM\ncow7x5ghz7hzjBkaG7eHhszMMudCYGaWudwKQfOv8DA95Rh3jjFDnnHnGDM0MO6s5gjMzOx0uR0R\nmJnZGC4EZmaZy6YQSLpK0p8lHZB0U6v70wySLpb0oKQnJD0uaV1qv0DS/ZKeSn/nt7qvjSapTdIf\nJd2b9jsl7Un5/pmkxl7bbxqQNE/SDklPStov6Z2Z5Ppz6f29T9Idkiply7ekH0o6KmlfTdu4uVVh\nc4q9X9IV9b5eFoVAUhvQA1wNLAGuk7Sktb1qiheAz0fEEmAZ8KkU501Ab0QsBnrTftmsA/bX7G8C\nvhMRlwD/Arpb0qvm+h5wX0S8BbiUIv5S51rSIuAzwJUR8TagDfg45cv3j4GrxrRNlNurgcXp9kng\n5npfLItCALwDOBARz0TE88CdwMoW96nhIuJwRPSl7RMUHwyLKGLdlh62DfhIa3rYHJKqwAeBW9O+\ngOXAjvSQMsY8F3gvcBtARDwfEccoea6TWcBsSbOAc4HDlCzfEfEb4J9jmifK7UrgJ1H4AzBP0kX1\nvF4uhWAR8GzN/mBqKy1JHcDlwB5gYUQcTncdARa2qFvN8l3gi8CptH8hcCwiXkj7Zcx3JzAE/CgN\nid0q6TxKnuuIOAR8CxigKADHgUcpf75h4txO+vMtl0KQFUnnAzuBz0bEv2vvi+J84dKcMyzpQ8DR\niHi01X2ZYrOAK4CbI+Jy4L+MGQYqW64B0rj4SopC+DrgPE4fQim9Ruc2l0JwCLi4Zr+a2kpHUjtF\nEbg9Iu5Kzf8YPVRMf4+2qn9N8G7gw5IOUgz5LacYO5+Xhg6gnPkeBAYjYk/a30FRGMqca4Au4K8R\nMRQRJ4G7KN4DZc83TJzbSX++5VIIHgYWpzMLXkMxubSrxX1quDQ2fhuwPyK+XXPXLuCGtH0DcM9U\n961ZIuLLEVGNiA6KvD4QEdcDDwIfSw8rVcwAEXEEeFbSm1PTB4AnKHGukwFgmaRz0/t9NO5S5zuZ\nKLe7gE+ks4eWAcdrhpDOTkRkcQOuAf4CPA18pdX9aVKM76E4XOwH9qbbNRRj5r3AU8CvgAta3dcm\nxf8+4N60/UbgIeAA8HPgnFb3rwnxXgY8kvJ9NzA/h1wDXweeBPYBPwXOKVu+gTso5kBOUhz9dU+U\nW0AUZ0U+DTxGcUZVXa/nJSbMzDKXy9CQmZlNwIXAzCxzLgRmZplzITAzy5wLgZlZ5lwIzMaQ9KKk\nvTW3hi3cJqmjdkVJs+lg1pkfYpad/0XEZa3uhNlU8RGB2VmSdFDSNyU9JukhSZek9g5JD6S14Hsl\nvT61L5T0C0l/Srd3pX/VJumWtKb+LyXNbllQZrgQmI1n9pihoWtr7jseEUuBH1CsegrwfWBbRLwd\nuB3YnNo3A7+OiEsp1gF6PLUvBnoi4q3AMeCjTY7H7BX5l8VmY0j6T0ScP077QWB5RDyTFvc7EhEX\nShoGLoqIk6n9cEQskDQEVCPiuZr/0QHcH8XFRZD0JaA9Ir7R/MjMxucjArP6xATb9XiuZvtFPFdn\nLeZCYFafa2v+/j5t/45i5VOA64Hfpu1e4EZ46ZrKc6eqk2b18DcRs9PNlrS3Zv++iBg9hXS+pH6K\nb/XXpbZPU1wp7AsUVw1bndrXAVsldVN887+RYkVJs2nFcwRmZynNEVwZEcOt7otZI3loyMwscz4i\nMDPLnI8IzMwy50JgZpY5FwIzs8y5EJiZZc6FwMwsc/8HqS73+esEApoAAAAASUVORK5CYII=\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYIAAAEWCAYAAABrDZDcAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nnbESTghYAAAAAElFTkSuQmCC\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYIAAAEWCAYAAABrDZDcAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nAAAAAElFTkSuQmCC\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYIAAAEWCAYAAABrDZDcAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nf5KqPyx7hyHiAAAAAElFTkSuQmCC\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    }
  ]
}