ReutersNLP.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "name": "ReutersNLP.ipynb",
      "provenance": [],
      "collapsed_sections": [
        "A6xT4uoraRqt"
      ],
      "authorship_tag": "ABX9TyMeEAV+QLR+2nR3vKEZ6Syg",
      "include_colab_link": true
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    },
    "language_info": {
      "name": "python"
    }
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "view-in-github",
        "colab_type": "text"
      },
      "source": [
        "<a href=\"https://colab.research.google.com/github/ashwani-09/esg_topicmodelling/blob/master/ReutersNLP.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "KerasNLP - Reuters\n",
        "=================\n",
        "First Iteration\n",
        "---------------\n",
        " 1. Build a simple model with entire data from Reuters dataset\n",
        " 2. Model architechture\n",
        "  ```\n",
        "  InputLayer()\n",
        "  Dense(512, \"relu\")\n",
        "  Dropout(0.5)\n",
        "  Dense(46, \"softmax\") --> Ouput\n",
        "  ```\n",
        " 3. Metrics\n",
        "  ```\n",
        "  loss=\"categorical_crossentropy\",\n",
        "  optimizer=\"adam\",\n",
        "  metrics=[\"categorical_accuracy\"]\n",
        "  ```\n",
        " 4. Result\n",
        "  ```\n",
        "  score = (14.71, 0.344)\n",
        "  ```\n",
        "\n",
        "Second Iteration\n",
        "----------------\n",
        " 1. Pick only 512 most frequent words\n",
        " 2. Model architechture\n",
        "  ```\n",
        "  Model: \"sequential\"\n",
        "  _________________________________________________________________\n",
        " Layer (type)                Output Shape              Param #   \n",
        "=================================================================\n",
        " embedding (Embedding)       (None, 512, 512)          262656    \n",
        "                                                                 \n",
        " dropout (Dropout)           (None, 512, 512)          0         \n",
        "                                                                 \n",
        " global_average_pooling1d (G  (None, 512)              0         \n",
        " lobalAveragePooling1D)                                          \n",
        "                                                                 \n",
        " dropout_1 (Dropout)         (None, 512)               0         \n",
        "                                                                 \n",
        " dense (Dense)               (None, 46)                23598     \n",
        "                                                                 \n",
        "  =================================================================\n",
        " ```\n",
        " 3. Results\n",
        "  ```\n",
        "  score = (1.3, 0.72)\n",
        "  ```"
      ],
      "metadata": {
        "id": "TLVYR23QcWPo"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "import numpy as np\n",
        "import sys, os\n",
        "\n",
        "import tensorflow as tf\n",
        "from tensorflow import keras\n",
        "from tensorflow.keras import layers\n",
        "\n",
        "from sklearn.metrics import ConfusionMatrixDisplay, confusion_matrix\n",
        "from pprint import pprint\n",
        "import plotly.graph_objects as go\n",
        "import matplotlib.pyplot as plt"
      ],
      "metadata": {
        "id": "LqiZlQFFYL_4"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "**Prepare data**\n",
        " - Pad unequal sequences\n",
        " - Convert labels to one-hot encoded variables for multi-class categorical classification\n",
        " - Build an index-word map to obtain original sentences"
      ],
      "metadata": {
        "id": "Ut4OLxY7YOp4"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "word_index = keras.datasets.reuters.get_word_index()"
      ],
      "metadata": {
        "id": "xNnJIwwtCh6k"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "index_word = {}\n",
        "for word, index in word_index.items():\n",
        "  index_word[index] = word"
      ],
      "metadata": {
        "id": "84pzy-CaCot7"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "(x_train, y_train), (x_test, y_test) = keras.datasets.reuters.load_data(num_words=512, seed=113)"
      ],
      "metadata": {
        "id": "Kp7fZ1_kBXdJ"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "training_data_corpus = np.hstack(x_train)\n",
        "print(\"Number of unique words in training data: \", len(np.unique(training_data_corpus)))\n",
        "print(\"Min index in training data: \", min(training_data_corpus))\n",
        "print(\"Max index in training data: \", max(training_data_corpus))"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "o-RfIAoe-LUT",
        "outputId": "2db00541-ca72-40c0-d08c-d8ccdc6bce02"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Number of unique words in training data:  510\n",
            "Min index in training data:  1\n",
            "Max index in training data:  511\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "\" \".join([index_word[index] for index in x_train[0] if index != 0])"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 92
        },
        "id": "J8cRfHaVCsuQ",
        "outputId": "608903ee-6dcf-4036-dd2d-47890d747a0b"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "application/vnd.google.colaboratory.intrinsic+json": {
              "type": "string"
            },
            "text/plain": [
              "'the of of mln loss for plc said at only ended said of could 1 traders now april 0 a after said from 1985 and from foreign 000 april 0 prices its account year a but in this mln of an states earlier and rise and revs vs 000 its 16 vs 000 a but 3 of of several and shareholders and dividend vs 000 its all 4 vs 000 1 mln agreed of april 0 are 2 states will billion total and against 000 pct dlrs'"
            ]
          },
          "metadata": {},
          "execution_count": 104
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "To remove most frequent and least frequent words from training data, we can simply remove the lowest and highest indices from training data."
      ],
      "metadata": {
        "id": "XaJkwHjyBivA"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "MOST_COMMON_WORDS_TO_REMOVE = 50\n",
        "MOST_COMMON_WORDS_TO_KEEP = 128\n",
        "unique_index = np.unique(training_data_corpus)\n",
        "sorted_unique_index = sorted(unique_index)\n",
        "keep_index = sorted_unique_index[MOST_COMMON_WORDS_TO_REMOVE:MOST_COMMON_WORDS_TO_KEEP]\n",
        "print(min(keep_index), max(keep_index))"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "bTWCiz4RI8mw",
        "outputId": "8c46a7d7-cad9-4755-dd83-1b23716b9cbe"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "52 129\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "%%time\n",
        "# x_train = [[index for index in row if index in keep_index] for row in x_train]"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "Ke_fzXhaKeY2",
        "outputId": "5bac1454-0ce4-468e-b303-f273ff22cd73"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "CPU times: user 3 µs, sys: 0 ns, total: 3 µs\n",
            "Wall time: 8.58 µs\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "print(x_train[0])"
      ],
      "metadata": {
        "id": "rmzOghIFcj2H",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "1403eb50-7bcb-4d18-bd49-e1d6bff73dd9"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "[111, 90, 67, 89, 102, 124, 90, 67, 84, 83, 90, 67, 52, 109]\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "def plot_lengths(data):\n",
        "  data_len = [len(x) for x in data]\n",
        "\n",
        "  fig = go.Figure(data=[go.Histogram(x=data_len)])\n",
        "  fig.show()"
      ],
      "metadata": {
        "id": "pud-cqcBBjr8"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "plot_lengths(x_train)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 542
        },
        "id": "jS6jypH6672U",
        "outputId": "37d4526f-6d0e-488e-b5e9-06fc7f44e851"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/html": [
              "<html>\n",
              "<head><meta charset=\"utf-8\" /></head>\n",
              "<body>\n",
              "    <div>            <script src=\"https://cdnjs.cloudflare.com/ajax/libs/mathjax/2.7.5/MathJax.js?config=TeX-AMS-MML_SVG\"></script><script type=\"text/javascript\">if (window.MathJax) {MathJax.Hub.Config({SVG: {font: \"STIX-Web\"}});}</script>                <script type=\"text/javascript\">window.PlotlyConfig = {MathJaxConfig: 'local'};</script>\n",
              "        <script src=\"https://cdn.plot.ly/plotly-2.8.3.min.js\"></script>                <div id=\"f28cd68b-df07-4f3d-ad76-c8570767d45b\" class=\"plotly-graph-div\" style=\"height:525px; width:100%;\"></div>            <script type=\"text/javascript\">                                    window.PLOTLYENV=window.PLOTLYENV || {};                                    if (document.getElementById(\"f28cd68b-df07-4f3d-ad76-c8570767d45b\")) {                    Plotly.newPlot(                        \"f28cd68b-df07-4f3d-ad76-c8570767d45b\",                        [{\"x\":[87,56,139,224,101,116,100,100,82,106,31,59,65,316,527,76,114,17,91,77,231,108,83,29,95,110,23,373,114,354,133,222,571,155,83,208,170,269,74,19,23,78,21,377,104,299,89,56,94,139,118,36,137,107,83,66,70,112,88,51,83,123,17,185,84,52,102,73,106,486,107,82,263,172,491,190,143,62,26,88,114,38,85,112,793,104,25,21,101,28,25,81,135,73,62,18,90,266,91,64,409,92,319,166,85,200,152,474,60,58,584,71,273,51,255,28,171,143,91,46,110,77,162,24,52,175,61,95,97,259,112,73,94,17,650,23,88,309,105,89,41,314,54,374,67,160,75,207,373,84,87,74,98,136,256,95,50,108,52,100,233,86,68,47,73,158,113,23,100,80,202,104,122,101,95,118,253,192,80,183,29,604,41,103,63,147,472,58,23,37,98,73,122,812,51,179,56,60,60,25,212,34,406,122,27,86,83,43,208,26,159,65,48,335,701,314,333,68,263,124,337,180,67,81,41,92,96,185,51,75,196,54,47,48,44,166,74,111,96,167,88,100,104,36,139,35,25,17,48,88,68,36,63,115,152,82,100,181,88,109,124,234,57,199,49,131,188,192,122,290,153,297,50,71,145,185,110,93,79,116,685,28,94,260,100,88,426,24,304,28,390,224,43,48,49,17,193,158,192,169,132,132,75,94,146,41,508,61,60,121,310,232,49,49,274,297,66,142,43,17,38,245,159,57,92,149,30,67,98,16,95,481,16,16,144,178,90,56,104,303,56,222,42,222,105,87,510,30,59,81,111,82,242,26,65,67,109,83,35,100,82,106,109,58,107,65,34,41,76,57,102,748,117,23,203,78,496,139,641,75,496,524,110,17,207,34,103,96,91,170,42,100,29,807,86,58,827,116,407,174,90,100,171,59,105,255,175,171,64,17,646,103,74,226,154,18,188,96,80,719,39,70,138,392,117,70,358,188,85,83,126,93,93,17,450,34,103,200,592,35,157,82,53,67,17,27,191,222,40,31,25,133,80,226,822,16,171,204,154,81,217,82,707,100,602,48,78,44,395,44,79,53,17,55,198,70,107,161,189,79,50,82,275,64,57,41,226,68,94,118,58,75,20,173,153,199,104,52,71,107,170,79,158,202,221,200,92,188,218,51,129,75,21,17,130,134,384,61,169,25,60,293,57,105,99,112,205,593,205,239,98,94,629,397,65,220,200,132,51,269,309,204,91,54,26,163,519,91,81,208,190,73,99,96,103,75,37,45,17,874,74,107,30,28,17,30,234,16,66,233,143,373,71,92,88,311,113,674,56,40,37,107,242,171,104,258,500,75,87,90,110,126,53,17,304,23,90,57,51,88,207,100,434,73,39,17,34,199,302,136,65,119,36,310,70,136,156,229,170,136,52,84,285,158,34,16,93,98,115,197,88,175,55,304,100,74,90,109,16,66,90,88,72,102,188,140,190,72,211,103,28,85,74,55,65,90,736,66,45,133,87,216,175,62,43,576,501,66,771,703,433,73,126,95,113,90,75,91,192,120,98,328,41,54,41,43,65,248,40,98,18,149,74,49,51,93,104,99,61,62,82,81,71,83,346,610,149,130,71,199,25,1032,87,368,143,57,59,166,217,505,32,37,70,643,24,58,49,62,287,52,693,17,722,282,192,24,58,169,89,123,564,269,109,44,173,229,184,97,583,100,41,47,58,377,68,86,17,50,177,102,600,73,192,67,79,53,78,56,187,17,17,115,94,79,67,238,85,313,200,17,95,130,143,63,244,267,18,183,178,191,78,86,40,97,232,93,83,25,90,91,36,40,150,93,204,68,78,27,316,154,383,110,80,105,43,66,105,110,60,81,160,17,56,53,149,83,177,144,63,92,244,90,39,109,455,80,17,99,189,113,74,130,64,708,105,250,68,43,33,92,90,140,746,17,91,65,146,256,485,102,17,55,51,89,58,92,93,100,225,85,113,143,86,52,113,302,101,136,296,143,88,102,171,156,251,412,143,290,21,310,63,158,198,90,77,374,17,34,186,90,50,56,90,120,130,45,96,108,185,44,187,59,17,67,269,114,234,57,94,734,102,93,240,402,108,82,17,50,65,18,77,49,65,186,153,26,113,363,49,295,81,96,137,279,308,453,29,101,138,91,253,229,84,899,126,92,192,49,184,49,19,177,241,171,22,112,472,79,49,101,41,25,69,41,142,232,52,221,87,44,35,16,306,108,41,102,629,27,254,95,65,261,25,287,416,94,63,27,133,51,626,17,442,25,321,99,46,34,101,75,809,311,743,115,228,287,171,92,57,43,53,30,98,287,166,98,91,17,81,165,140,99,194,89,115,117,48,162,484,197,17,224,89,140,73,115,78,252,23,42,104,92,93,99,72,129,197,261,193,484,137,63,96,47,19,170,85,57,303,153,113,79,130,81,94,48,103,84,43,94,24,194,125,97,61,206,17,32,768,25,90,94,59,61,65,363,32,291,30,175,290,183,128,56,274,76,52,30,96,281,429,84,266,252,84,92,55,98,32,518,60,249,70,573,55,58,107,320,53,331,87,245,122,95,392,92,431,60,651,206,36,94,516,444,149,405,102,333,49,216,631,17,118,149,19,69,42,53,114,103,86,96,470,138,39,96,86,126,171,112,168,73,89,43,141,60,665,119,202,546,36,85,55,78,57,377,83,17,62,111,86,110,89,288,52,65,72,103,83,40,240,403,87,110,333,21,249,69,17,62,98,203,51,375,38,491,75,99,76,319,187,48,78,80,159,293,205,70,188,25,102,56,87,281,134,280,82,302,229,59,199,182,23,63,121,282,58,752,418,76,52,28,162,110,81,100,159,26,242,408,116,56,139,162,29,217,89,113,36,167,87,90,249,160,35,72,35,58,97,23,91,191,105,102,65,84,367,97,51,108,48,237,621,81,89,369,320,163,103,164,98,94,86,30,422,279,48,233,35,145,78,246,79,86,91,16,148,84,77,261,65,80,32,808,29,196,17,58,201,43,192,79,127,113,242,42,46,184,60,82,99,103,87,47,158,176,42,97,30,81,225,40,223,186,152,504,95,100,88,108,174,109,164,94,59,193,128,97,297,210,169,27,78,17,33,89,102,249,88,62,33,144,200,174,92,214,97,103,15,192,295,89,79,45,52,210,121,329,52,115,152,88,136,17,97,227,61,329,95,260,190,241,99,40,55,167,401,196,154,321,63,183,67,73,67,206,151,72,75,74,52,61,141,93,101,67,224,31,59,657,43,526,110,114,536,105,280,215,85,80,27,78,68,57,105,176,17,17,74,86,41,136,91,107,82,245,51,82,62,79,62,174,391,65,262,26,63,172,75,153,165,266,27,55,49,59,266,185,68,42,103,267,69,57,17,551,289,46,111,103,88,87,249,134,596,50,226,43,130,52,304,63,212,17,52,91,165,73,114,122,204,28,203,102,193,95,76,413,43,67,58,109,161,68,147,35,17,116,59,324,104,59,248,68,94,193,17,80,130,371,44,148,110,74,76,393,79,17,482,96,294,57,58,144,21,81,137,108,80,188,24,84,219,108,170,80,48,176,47,19,274,433,26,81,372,68,26,83,109,92,154,192,17,93,74,67,96,27,85,178,634,44,163,77,129,65,89,69,164,32,533,199,97,255,195,132,101,39,117,115,62,128,102,312,71,141,93,356,35,167,90,305,324,53,143,467,160,383,310,438,282,381,204,85,24,111,90,41,51,42,108,282,244,127,904,48,816,57,108,83,99,193,78,222,301,85,40,659,50,242,27,48,144,88,123,163,339,199,386,416,43,89,35,49,211,82,21,248,306,164,99,279,612,601,74,383,287,568,102,64,65,139,81,91,231,126,82,70,153,69,165,103,354,89,202,186,422,94,220,92,241,145,70,62,68,61,191,116,54,16,17,95,89,38,180,174,41,106,215,55,38,23,82,172,226,16,53,229,69,96,42,82,27,66,299,187,284,55,33,84,118,200,46,492,28,159,249,66,190,159,150,115,696,95,83,121,154,77,17,91,198,213,53,284,51,269,754,75,165,459,144,262,522,213,65,49,80,17,38,157,292,147,342,52,336,54,82,76,100,71,27,16,593,91,105,37,17,133,25,67,158,47,186,175,71,774,409,206,172,77,66,386,89,163,252,78,75,67,111,101,217,77,181,108,98,42,36,61,262,67,607,285,17,430,169,107,79,51,71,95,109,257,50,42,138,91,54,90,17,261,236,111,82,99,88,275,42,37,51,79,132,32,75,52,265,53,236,165,50,335,103,56,17,52,141,107,156,93,280,342,93,79,214,51,28,696,302,69,105,71,17,60,215,584,49,103,86,618,37,17,636,54,17,194,17,45,111,146,183,94,316,75,73,82,314,28,64,483,288,303,35,156,261,362,245,199,122,369,553,17,104,45,266,72,99,234,50,97,35,206,117,492,17,280,67,81,66,17,83,102,101,37,52,58,17,75,25,75,196,354,138,134,693,196,94,91,121,77,41,26,99,86,91,87,74,17,72,98,497,471,34,29,206,50,21,76,54,50,26,87,24,417,100,24,672,225,67,352,280,143,121,88,49,167,84,131,219,47,96,90,71,48,542,57,23,217,98,88,84,88,77,53,254,2246,84,35,106,66,394,81,61,52,83,206,87,83,50,29,253,39,83,43,77,430,137,72,21,201,117,220,88,97,348,167,178,16,79,109,64,105,94,19,57,57,162,94,17,62,72,237,227,53,52,131,59,96,249,42,86,54,82,122,36,52,98,32,188,73,58,313,49,85,253,115,23,121,107,162,96,76,373,58,17,83,90,36,62,109,612,85,23,17,92,211,104,97,403,95,213,58,55,228,67,67,21,96,94,414,17,90,582,51,43,23,129,161,69,192,62,54,78,337,57,20,109,57,86,479,49,54,219,102,134,23,54,140,265,292,129,155,44,131,487,40,83,93,39,88,116,399,177,510,52,339,207,99,52,143,177,125,113,74,43,111,79,63,103,282,282,165,77,30,23,68,258,45,70,286,25,104,280,84,436,115,52,192,69,47,207,162,275,116,395,49,108,86,49,223,71,46,729,93,36,26,83,92,168,415,181,272,69,77,259,23,75,172,148,17,143,34,675,192,27,41,42,25,236,165,18,416,83,165,99,61,47,49,546,23,52,280,112,105,245,97,363,254,400,34,125,152,50,17,64,123,56,90,75,302,252,70,75,165,134,78,69,85,108,80,26,445,66,39,92,175,366,51,98,273,209,60,64,101,298,60,86,81,214,26,26,36,105,55,131,94,94,22,47,180,186,200,86,364,113,206,27,87,31,234,301,84,111,46,65,501,67,50,80,104,145,60,50,48,117,60,43,51,190,69,93,92,163,700,21,217,17,187,121,474,83,98,63,130,178,83,43,127,299,147,96,50,75,215,70,181,101,96,714,37,94,323,385,53,110,68,30,284,95,52,34,244,98,120,116,103,624,101,103,102,325,606,17,364,161,109,55,25,68,67,41,91,77,66,548,87,99,62,48,77,391,79,175,25,163,158,80,541,263,418,69,77,107,100,83,414,20,28,135,50,78,400,70,235,80,67,127,650,216,106,73,510,42,197,58,167,65,158,101,99,240,193,433,140,61,37,54,86,108,209,17,255,117,53,91,111,93,107,68,302,202,78,315,94,83,43,84,341,113,138,53,85,36,35,423,102,59,76,85,208,43,67,452,63,29,441,132,37,62,115,644,94,43,64,111,356,58,108,479,52,17,72,275,106,224,83,210,79,422,35,95,378,20,31,53,453,86,103,54,468,27,54,271,340,32,17,72,31,85,105,117,16,22,278,82,447,90,83,600,42,42,79,106,128,47,24,96,79,92,38,17,212,87,110,43,316,40,178,66,97,102,42,21,74,78,115,97,112,91,116,42,190,57,302,270,82,157,254,74,94,196,63,145,63,357,36,344,45,81,62,72,166,25,175,74,155,108,28,461,182,353,121,188,306,36,61,109,128,66,21,61,118,136,71,17,56,38,49,18,110,81,61,62,122,116,52,23,261,196,50,89,68,108,54,110,59,311,49,86,85,319,311,217,111,51,171,520,47,55,540,459,220,41,102,196,213,158,103,90,123,222,431,69,138,218,32,32,107,57,122,108,102,369,216,223,18,74,268,502,102,123,87,94,45,188,168,48,73,34,137,465,214,59,71,76,51,158,109,57,57,112,17,329,217,101,77,298,58,252,31,254,137,46,67,246,160,71,194,461,144,76,44,54,175,63,51,49,82,84,207,44,48,35,289,190,19,474,106,20,218,84,628,21,90,145,89,52,64,43,19,49,48,303,63,138,39,162,94,57,27,103,18,105,556,36,109,561,17,246,809,313,80,92,177,206,81,58,61,129,315,191,72,54,130,130,54,56,94,95,624,54,166,62,21,317,151,74,77,78,173,144,152,550,73,92,96,744,30,233,65,96,125,133,77,106,83,31,177,54,216,244,51,513,153,55,44,36,74,70,25,17,17,100,108,17,135,76,719,60,87,111,54,46,21,183,78,27,54,96,120,344,355,108,194,17,59,75,585,70,586,118,627,91,79,68,226,155,250,144,79,89,82,468,58,193,76,124,259,63,67,72,261,26,82,300,403,41,17,58,205,123,85,77,68,89,61,67,511,113,82,149,83,36,101,81,121,75,17,65,61,61,45,204,67,86,144,120,40,83,227,70,27,188,174,171,66,70,279,61,255,102,248,94,69,58,123,17,125,103,96,196,110,48,27,59,56,80,298,40,70,559,577,103,76,246,171,68,83,27,157,264,192,17,93,73,47,362,78,212,97,66,192,72,90,506,187,531,81,127,45,75,99,91,335,17,57,33,80,50,268,240,93,359,474,93,200,117,99,92,93,79,85,78,421,324,208,622,363,299,122,282,170,57,99,483,75,109,331,106,34,58,77,86,108,172,129,112,35,382,768,91,146,49,96,121,84,17,22,184,71,53,218,105,66,82,38,23,17,170,139,82,99,254,195,46,55,28,114,24,94,91,78,23,68,54,41,396,72,50,73,85,141,91,23,298,190,55,74,91,96,20,390,81,711,140,153,1053,310,74,65,391,61,37,49,79,36,25,17,476,90,197,49,39,97,92,35,342,101,248,197,50,117,175,208,22,504,67,63,51,32,62,60,41,163,67,97,375,32,57,84,38,381,66,94,107,93,19,98,104,84,107,41,67,43,741,108,55,38,113,67,76,38,75,278,330,469,59,108,76,92,148,270,96,41,18,224,54,67,94,277,184,101,23,266,186,64,214,162,55,92,17,133,100,91,66,105,122,63,291,144,595,113,181,35,57,246,148,267,135,210,73,84,176,193,103,74,86,72,350,85,109,75,249,72,27,322,43,63,90,124,121,34,172,541,96,94,117,278,67,53,53,106,395,113,55,393,23,77,53,97,154,95,17,66,59,187,53,307,149,45,29,42,99,40,48,48,229,544,84,278,20,160,94,85,199,298,76,65,58,104,41,93,67,152,63,23,31,828,76,42,485,105,118,66,227,18,96,93,35,76,61,81,181,219,280,53,41,274,204,163,223,98,38,40,115,104,320,455,102,45,120,52,94,172,370,385,178,109,514,178,346,58,128,93,47,244,82,68,35,128,116,483,66,49,191,101,240,329,88,72,106,176,25,63,111,208,125,216,60,82,87,40,73,123,35,28,58,135,172,223,319,17,102,99,504,303,51,424,17,137,27,85,62,47,76,49,105,109,71,40,49,42,76,98,94,231,50,59,83,63,307,130,99,113,213,80,175,591,195,174,54,93,94,59,49,140,651,98,83,178,160,62,334,91,68,41,533,628,33,99,65,140,93,21,112,98,158,345,95,134,31,114,158,73,213,547,109,194,89,108,472,75,65,85,245,37,234,23,91,77,17,122,70,47,100,57,93,108,67,363,378,91,102,119,29,109,359,270,280,70,50,105,67,76,59,48,68,147,47,218,17,55,24,76,78,97,103,67,71,40,68,235,267,44,17,142,35,89,91,87,227,216,83,108,476,109,25,85,503,90,424,107,36,61,99,78,100,81,68,35,171,642,210,101,38,358,17,37,87,78,445,84,295,358,676,73,55,165,115,106,246,113,438,283,196,195,117,475,96,170,83,27,260,162,31,35,21,80,64,38,100,86,162,98,96,60,81,77,55,79,41,121,55,94,67,100,99,105,134,76,648,778,204,165,231,309,83,91,96,74,442,116,63,98,100,84,414,75,172,50,81,50,34,474,331,87,64,167,304,71,110,65,53,60,401,59,82,48,99,25,184,224,37,66,87,99,89,36,85,27,106,65,85,92,241,88,54,209,111,218,17,194,40,225,95,785,183,48,16,88,92,278,37,65,21,61,175,27,78,59,74,50,79,47,61,84,69,49,272,705,881,103,111,117,73,47,347,76,55,97,59,43,331,302,218,101,17,74,187,71,57,158,87,25,74,630,54,35,94,101,67,67,109,243,210,153,28,143,296,17,24,30,250,173,606,115,119,69,68,69,86,45,46,84,293,60,76,196,30,95,205,114,92,68,670,111,36,59,191,77,96,104,39,209,94,25,58,17,212,475,314,356,27,158,110,17,111,113,105,60,42,85,106,18,27,69,240,103,128,287,193,59,92,232,67,54,27,51,25,34,155,57,293,50,79,179,98,72,74,69,87,302,192,101,72,74,211,245,73,56,590,97,59,112,102,64,134,83,29,524,193,27,62,102,370,93,106,213,88,67,16,123,103,44,537,17,61,156,27,318,197,81,276,481,76,476,89,86,102,96,184,91,427,93,41,212,84,119,37,658,97,252,59,99,188,274,97,25,176,86,39,167,90,35,84,292,74,105,97,81,217,28,129,34,42,130,129,163,144,90,17,96,114,217,51,104,105,166,483,27,391,106,194,185,671,61,25,100,21,139,175,55,120,251,84,207,191,170,147,267,209,169,53,211,650,208,588,475,159,46,62,224,43,23,110,258,43,125,86,116,472,698,26,85,187,27,86,163,56,44,92,66,259,119,336,57,57,165,25,320,205,16,113,26,212,241,227,90,52,91,154,62,187,62,186,35,100,34,102,88,276,23,25,99,77,19,22,208,76,130,25,116,520,25,199,91,115,115,121,71,82,476,289,243,61,306,138,164,250,70,109,118,230,40,71,59,684,83,291,16,141,92,71,353,139,194,184,112,62,237,51,191,17,34,288,304,76,723,27,193,74,146,118,87,116,267,514,26,80,174,31,25,28,78,223,165,79,72,58,41,92,66,235,121,63,116,366,81,571,88,72,45,95,17,129,58,35,43,130,395,94,21,104,848,67,62,460,113,225,19,419,226,44,107,83,69,36,62,349,99,85,59,85,59,73,17,27,104,619,108,219,97,38,57,248,61,72,211,55,526,301,115,91,135,153,47,67,309,68,91,59,44,118,143,17,212,184,67,43,162,184,93,39,53,90,17,273,26,450,234,593,170,51,121,312,94,78,186,69,407,51,66,55,609,53,60,287,66,96,140,163,44,96,90,250,82,1035,23,83,109,189,49,31,49,84,369,31,65,584,104,149,85,188,92,450,37,86,48,51,30,71,362,184,161,17,203,17,151,68,350,85,271,366,27,79,65,381,111,38,49,60,706,73,76,65,160,67,311,261,153,259,89,101,250,345,276,57,43,594,37,105,124,172,225,110,200,17,82,111,224,699,268,351,89,171,100,152,358,88,70,112,75,115,51,106,443,49,73,81,246,736,43,100,98,167,157,46,77,72,93,78,92,313,337,106,24,89,61,66,25,60,79,66,73,364,92,369,198,678,64,76,21,90,348,381,71,75,59,150,70,83,83,112,105,84,92,80,32,58,42,211,49,74,36,181,127,183,78,94,264,93,217,156,113,110,17,293,40,124,74,45,98,97,96,84,41,185,67,72,151,224,161,167,68,93,24,256,162,113,47,82,140,16,180,58,149,204,114,85,162,64,56,180,133,95,67,254,189,119,452,170,104,123,412,50,78,78,82,196,354,21,81,341,530,855,29,77,146,58,816,247,355,39,60,73,62,78,53,155,184,64,18,58,437,17,573,106,174,40,103,181,84,627,59,210,104,458,50,85,331,119,178,59,95,96,60,79,27,407,61,167,100,48,322,63,134,93,119,85,81,59,28,87,102,154,67,159,47,17,17,128,99,91,85,84,252,57,107,302,140,29,135,20,122,49,105,16,208,101,107,336,142,314,172,481,49,220,51,51,718,368,367,101,186,443,17,103,70,411,25,256,35,370,61,115,54,60,67,76,128,23,120,86,82,49,361,51,283,303,39,101,94,128,81,52,71,81,88,275,70,53,101,94,76,17,60,45,33,102,114,45,84,224,190,105,73,64,47,264,283,177,114,266,25,108,93,62,245,577,19,100,77,195,136,573,74,110,85,139,39,373,122,112,203,114,430,195,112,158,103,401,25,60,219,115,51,37,57,413,130,150,75,84,212,46,47,52,108,97,76,51,86,454,92,120,47,356,59,66,216,262,90,78,130,45,55,506,29,59,73,180,112,198,116,98,66,85,93,82,368,31,243,129,187,309,95,260,421,222,57,41,799,45,230,59,148,170,92,174,51,58,35,130,54,180,290,291,134,95,54,60,685,71,149,73,65,85,100,63,124,86,189,53,111,210,43,63,122,211,99,226,44,104,30,73,95,53,104,173,93,186,288,150,52,98,133,280,292,124,83,123,77,88,62,476,105,178,130,96,42,177,92,21,102,93,60,172,25,17,101,64,203,93,239,164,112,97,45,17,141,242,27,156,53,67,167,180,839,39,48,31,208,198,270,69,163,78,137,309,73,91,17,374,133,197,456,392,50,73,271,48,139,41,68,70,93,48,80,27,25,94,140,86,101,144,297,59,380,84,57,65,126,131,36,40,404,172,271,82,73,748,95,81,64,44,64,87,598,47,94,17,225,50,69,106,86,85,48,92,78,70,113,64,29,110,118,49,104,128,101,102,146,178,79,22,72,17,88,71,99,290,55,78,218,72,357,99,93,107,114,23,216,42,327,206,44,388,69,60,397,67,49,104,446,32,215,101,134,175,42,63,172,92,87,297,75,438,96,78,89,148,164,131,18,169,59,55,25,54,516,62,298,64,296,107,101,168,53,501,17,449,106,46,811,47,53,141,73,48,516,184,92,69,560,62,80,112,309,107,310,306,94,119,51,108,64,91,303,270,64,55,86,55,296,59,25,135,52,749,97,883,83,222,243,253,101,345,39,97,110,81,152,99,205,30,17,95,275,63,27,94,19,109,42,118,347,225,100,289,121,234,30,420,712,394,50,248,58,171,26,99,183,243,54,300,123,53,112,292,143,107,78,66,115,173,195,229,98,148,17,105,108,73,126,286,84,190,96,50,34,159,318,374,78,165,552,194,37,696,152,320,162,99,712,127,297,579,55,130,86,138,69,130,306,39,61,78,587,16,200,44,56,49,186,200,404,179,125,43,282,164,24,112,155,97,17,73,303,81,38,289,40,30,319,15,179,90,206,236,66,531,59,450,16,199,307,971,188,53,264,47,26,46,845,135,61,28,309,17,57,767,331,99,56,641,108,71,156,99,164,19,300,56,19,84,53,81,71,74,69,84,124,51,53,21,69,180,49,39,66,55,53,640,19,48,47,219,80,28,98,105,442,98,114,144,188,18,68,183,103,741,186,80,182,74,63,56,199,350,77,100,52,64,60,61,183,66,152,25,26,86,311,36,67,224,42,86,129,55,510,27,71,135,86,17,45,19,113,92,103,358,104,195,629,34,98,445,186,332,118,280,55,97,69,54,28,39,91,60,257,17,103,82,54,67,44,122,315,139,262,55,108,101,135,204,82,89,19,126,98,40,125,59,255,153,48,68,101,19,196,94,88,22,112,166,59,165,378,57,67,90,357,91,103,80,175,192,23,90,96,77,172,144,87,711,129,55,224,619,66,188,101,53,112,23,44,80,19,101,110,134,255,175,69,83,55,101,91,56,148,94,316,69,56,50,665,88,98,53,446,229,247,77,347,85,767,77,212,96,30,575,52,142,194,205,110,231,96,60,58,34,75,554,132,86,101,83,56,269,184,33,89,29,96,74,25,238,56,88,99,75,47,185,31,119,94,255,214,84,112,103,69,167,17,138,160,477,33,91,57,200,374,855,211,166,139,80,134,95,592,45,83,100,51,94,90,325,27,61,17,87,84,113,699,447,224,707,277,191,53,57,52,45,30,364,61,61,38,71,527,105,52,36,231,55,146,65,56,168,25,36,123,83,108,400,38,73,329,444,380,34,92,50,26,631,161,276,682,92,17,53,90,127,73,501,685,115,169,434,380,99,185,68,79,149,16,252,41,90,29,122,81,66,120,91,93,86,50,160,116,81,333,91,48,101,46,38,17,108,82,23,371,132,156,91,137,19,723,581,368,152,49,84,146,61,95,94,101,100,175,222,17,66,163,72,67,68,83,61,71,166,17,293,159,106,153,29,219,129,60,19,32,52,402,39,27,129,57,66,94,52,85,234,71,95,87,72,82,104,54,259,46,95,126,49,118,96,40,457,44,34,208,49,141,96,225,66,23,16,276,237,103,65,207,153,228,150,54,203,263,221,52,19,36,106,21,77,16,28,109,76,17,109,34,49,390,68,103,508,96,139,60,69,284,19,82,168,51,152,94,103,24,32,184,122,29,242,173,80,69,132,45,91,34,98,206,113,289,62,160,90,178,701,37,43,197,98,97,264,80,394,132,126,116,53,25,123,197,66,46,28,100,105,184,44,73,176,101,45,88,38,25,86,98,101,34,174,276,90,109,239,400,94,111,91,380,39,140,69,79,72,126,598,34,58,73,48,113,17,98,17,55,46,323,224,60,22,163,117,258,42,186,63,116,116,68,17,299,90,93,63,32,42,77,165,107,37,94,102,80,95,67,448,105,61,163,110,144,93,109,317,17,172,62,53,116,103,55,149,71,142,31,209,228,87,22,52,95,99,357,56,185,172,336,98,152,659,83,53,174,227,16,98,259,89,81,235,93,75,46,504,63,47,67,53,164,184,107,160,71,323,51,732,86,108,438,86,287,516,174,69,208,532,78,17,153,38,114,80,17,735,233,59,57,205,93,17,108,125,53,507,42,91,33,45,156,244,74,110,92,17,246,49,474,207,48,107,313,254,51,123,27,45,108,311,74,19,49,238,186,57,40,459,45,240,186,56,96,256,121,58,240,99,25,163,17,907,74,122,771,590,273,199,110,28,56,251,392,140,112,99,88,76,122,118,297,68,77,51,178,40,17,54,146,387,393,127,241,59,202,32,77,71,180,55,104,199,195,607,17,110,32,33,99,30,108,85,203,104,475,29,108,58,72,116,60,68,47,158,109,41,17,201,40,17,136,73,907,40,220,111,47,203,113,303,75,202,78,264,166,204,29,172,45,507,119,81,52,114,70,35,203,203,53,206,80,194,224,50,78,248,146,184,94,87,140,73,50,77,69,215,299,122,168,98,56,87,308,108,122,66,428,195,63,193,94,220,359,31,98,268,24,366,16,129,54,28,55,100,393,75,765,50,88,81,137,25,45,78,23,66,94,65,46,98,55,16,56,84,48,99,200,21,171,65,71,34,61,401,127,27,85,173,46,84,70,308,84,79,480,152,290,362,102,105,161,34,154,76,116,62,82,185,69,174,82,78,268,43,83,65,55,128,105,32,56,53,101,82,206,97,59,441,138,77,113,85,39,82,51,296,51,114,95,230,111,75,69,49,40,17,181,40,227,303,118,115,19,69,526,62,294,64,322,219,65,98,147,308,69,122,119,135,84,48,271,262,57,86,574,166,63,38,101,66,176,180,16,170,100,61,218,157,101,252,258,68,179,116,25,357,115,97,73,23,57,46,89,197,286,261,84,67,197,106,94,201,37,68,241,230,206,143,173,225,192,121,118,428,554,92,38,195,95,94,61,187,266,113,45,168,82,69,56,44,74,112,90,130,97,76,119,31,45,85,97,292,52,58,68,198,192,72,116,32,231,96,158,555,83,78,118,197,100,70,172,34,63,393,331,34,348,604,103,215,101,481,101,276,442,71,166,785,217,64,34,61,85,61,54,181,97,114,67,81,118,62,57,174,46,34,51,51,140,195,78,69,293,54,16,195,94,55,95,264,183,83,115,403,89,128,527,189,36,220,89,79,33,41,19,460,160,188,297,38,186,32,135,70,87,175,192,80,68,116,105,98,89,65,69,197,99,153,162,234,24,60,104,129,208,60,172,132,65,62,73,98,56,119,650,690,95,57,116,284,71,522,87,553,26,36,65,141,27,136,39,31,148,97,113,226,108,49,173,47,189,45,130,100,72,48,52,48,140,31,99,92,107,34,272,323,116,107,148,417,86,79,251,74,120,85,75,85,64,40,250,72,118,110,131,95,68,110,79,103,232,111,97,26,266,68,53,105,51,102,17,179,81,54,157,75,571,59,60,74,65,77,42,174,130,29,82,462,198,429,42,217,26,870,76,189,97,285,19,163,98,137,68,281,260,98,342,166,103,65,49,86,248,182,125,105,98,142,28,96,160,201,250,108,102,182,160,104,93,32,314,123,99,83,232,75,55,66,105,57,56,54,205,92,48,23,57,197,88,75,68,186,153,103,464,95,72,92,62,49,124,217,28,68,28,202,214,96,105,231,27,106,80,37,33,256,106,61,17,203,311,616,244,93,224,201,21,36,69,48,56,186,190,53,445,570,46,694,35,43,466,316,61,81,207,109,13,199,23,58,74,96,177,103,15,263,55,537,129,230,111,98,33,842,74,104,100,355,153,44,60,83,117,245,149,35,128,149,729,582,45,76,86,48,84,16,55,64,244,67,46,541,320,21,193,38,257,447,114,290,104,108,66,181,58,198,27,61,17,203,61,493,339,22,600,515,171,41,311,29,50,131,69,105,17,349,50,65,156,213,38,73,78,76,183,67,66,51,69,53,55,70,17,416,26,128,167,104,549,40,150,835,93,91,42,59,68,58,84,374,293,138,75,650,94,22,74,17,107,39,86,396,972,60,54,229,230,61,20,678,118,579,52,77,71,88,111,71,30,59,541,268,62,159,34,178,204,173,49,96,30,120,181,72,39,68,71,174,51,58,38,65,109,323,150,40,269,89,45,100,39,209,106,97,73,82,791,64,80,45,115,50,416,61,36,36,107,401,107,58,409,56,98,44,29,16,106,175,283,71,99,17,45,125,91,103,145,319,101,102,109,76,183,296,466,857,188,95,46,41,521,34,33,186,100,87,157,698,310,120,168,171,176,92,78,67,62,296,86,57,25,42,290,822,53,259,209,48,17,34,161,68,93,314,117,94,74,97,36,47,17,35,21,102,18,104,102,52,92,78,97,114,194,111,102,111,75,72,55,105,70,108,816,97,123,125,160,50,418,93,201,293,31,103,79,121,118,342,19,122,38,39,115,49,59,335,17,210,81,112,73,76,123,52,177,195,126,21,107,185,125,102,294,301,86,27,17,60,41,469,103,203,72,313,29,205,40,125,718,155,197,19,375,105,88,116,67,279,75,48,23,64,66,61,196,116,201,214,40,43,373,21,203,162,30,51,188,83,69,64,54,96,16,54,89,62,275,36,864,174,671,107,280,269,71,46,27,73,114,473,92,117,17,92,63,71,17,78,110,95,102,356,18,110,178,22,183,429,42,33,257,190,107,59,295,37,29,76,326,63,112,202,167,219,108,17,97,34,372,569,205,24,218,312,106,73,100,340,27,96,156,42,163,59,36,90,216,55,42,36,73,530,106,108,32,80,21,37,311,23,32,41,449,26,30,130,143,99,35,84,137,535,79,106,42,53,82,41,37,209,17,122,65,27,123,55,110,129,36,25,111,76,33,232,70,664,197,71,40,17,135,97,42,52,192,214,21,55,134,92,33,548,32,216,66,30,365,101,177,148,118,16,48,105,114,31,103,207,171,593,417,292,17,150,459,41,200,54,29,180,283,367,78,66,119,49,71,115,85,17,713,78,173,130,408,53,127,119,279,49,98,186,163,209,85,141,80,53,48,279,87,43,616,87,184,89,87,179,103,242,113,145,44,79,51,509,25,68,101,34,87,35,266,70,50,44,122,105,159,69,80,144,191,78,99,135,53,50,79,176,44,27,54,151,363,185,99,46,143,27,31,41,87,37,510,295,522,114,60,104,115,183,110,336,47,122,471,61,138,77,88,46,56,50,97,22,96,86,207,181,200,124,39,46,131,102,56,50,172,37,254,71,49,125,104,367,64,150,325,56,167,78,463,89,62,41,102,165,23,51,358,222,55,73,18,81,328,396,28,57,72,92,132,53,29,78,55,359,74,376,105,29,173,60,29,321,78,170,102,309,99,78,218,143,56,200,31,88,282,99,79,166,372,126,302,78,205,78,65,77,266,68,16,223,271,193,61,32,352,306,17,57,21,88,202,25,35,80,95,71,86,463,306,67,101,86,232,172,96,114,92,33,55,49,81,63,253,207,48,17,89,87,567,36,73,141,46,65,78,189,179,129,153,72,103,229,103,67,187,52,206,75,102,81,48,73,50,38,27,91,314,75,199,105,50,58,90,64,162,63,51,163,94,142,75,95,126,19,71,82,390,85,72,49,145,75,176,64,195,129,171,231,324,71,52,85,77,111,83,104,97,102,202,255,85,364,49,92,44,466,82,100,630,145,66,71,27,123,116,21,204,244,108,58,61,193,101,72,648,97,221,136,219,44,113,62,31,496,137,791,179,37,130,118,73,187,98,499,87,82,35,84,88,45,101,119,167,24,177,115,193,125,96,73,72,38,71,293,82,327,37,596,125,69,165,25,59,89,218,186,70,293,295,80,58,80,199,131,292,283,56,339,111,160,54,65,117,25,594,144,83,27,98,77,111,86,198,84,116,202,94,80,519,70,180,100,21,74,411,143,100,170,298,115,699,47,478,614,99,103,108,103,74,76,159,85,48,69,93,311,17,25,808,97,135,279,85,80,144,64,259,706,111,18,297,17,25,93,108,62,89,100,176,188,77,58,162,89,137,55,66,26,29,171,97,38,99,26,110,180,124,127,348,44,91,92,20,60,55,67,196,175,76,92,52,124,73,112,200,104,53,51,297,41,133,237,563,141,76,62,49,17,75,77,106,38,28,151,127,85,573,51,102,213,297,110,86,155,343,121,298,283,174,131,181,71,70,89,72,81,69,358,331,338,58,39,84,92,73,280,108,77,205,93,260,326,85,96,73,190,102,17,94,69,209,141,100,437,396,263,178,153,98,17,103,35,93,86,98,42,100,70,83,476,120,110,608,70,68,107,380,72,266,39,146,100,333,124,87,78,110,65,70,38,551,56,158,171,84,697,150,78,64,224,167,51,84,251,172,26,51,106,428,143,418,188,36,207,116,615,285,70,201,95,81,144,149,114,65,121,242,137,148,104,168,195,88,168,148,863,78,46,103,105,94,51,306,52,172,105,155,386,54,178,274,150,105,187,156,152,85,92,182,32,207,352,39,94,82,389,19,79,79,53,81,146,628,118,56,292,60,34,253,66,18,35,138,104,111,27,165,89,65,172,189,54,223,292,17,117,97,63,81,225,16,125,75,57,162,83,102,48,49,588,17,108,133,287,338,367,138,92,84,71,291,17,111,54,106,114,157,101,119,63,248,93,239,117,454,239,88,93,198,64,53,69,25,90,123,198,110,111,115,116,75,90,405,17,124,99,69,93,63,60,123,114,55,119,226,250,57,132,56,119,171,45,124,32,115,64,95,200,85,40,146,40,16,45,175,626,224,208,80,106,71,46,93,303,182,202,52,101,68,100,173,210,166,282,396,31,166,113,25,42,57,17,23,75,753,889,116,115,83,50,75,108,373,66,51,167,17,37,42,202,480,75,56,187,60,340,112,124,367,46,118,88,60,148,88,309,77,247,44,49,25,94,63,38,274,54,89,772,59,282,58,488,65,55,64,180,27,98,81,72,78,222,106,144,63,108,57,17,18,219,70,87,220,112,54,47,114,259,132,114,188,72,78,494,72,167,198,196,119,111,93,98,126,136,381,81,84,151,67,65,53,81,99,81,60,40,44,64,20,91,87,95,26,267,56,72,116,59,102,307,16,478,33,28,49,217,161,165,194,53,112,65,91,83,17,120,222,74,41,47,37,76,66,149,106,613,17,138,37,782,34,162,89,17,450,24,63,131,146,306,77,45,54,102,115,167,139,114,17,43,84,111,239,132,82,68,72,130,31,154,165,108,102,155,39,23,94,73,78,58,56,114,78,145,79,94,71,100,129,754,186,91,56,88,150,113,75,401,135,32,229,118,144,41,92,63,153,95,87,153,40,98,121,244,71,57,58,56,81,92,92,97,94,154,163,319,272,17,66,93,67,94,183,21,680,55,316,75,176,145,77,54,17,99,65,48,113,88,259,26,129,47,70,230,64,59,116,23,460,71,70,480,65,474,235,129,302,18,75,83,915,111,36,157,36,101,563,212,109,80,172,62,85,109,79,59,112,73,73,190,115,1332,27,25,108,73,346,207,205,92,129,190,46,157,61,247,648,211,53,74,199,73,92,131,53,177,70,39,98,17,145,111,59,41,428,99,91,75,390,282,123,19,36,139,98,59,92,434,89,517,122,62,83,225,32,204,70,78,47,107,100,92,39,90,23,451,63,294,211,252,90,104,56,185,71,71,88,26,57,101,46,17,53,17,86,44,102,447,73,151,203,129,24,92,41,180,93,103,59,186,82,102,272,50,60,93,92,39,180,301,220,21,57,119,113,73,17,148,143,343,365,778,488,166,41,17,89,21,17,46,186,849,81,192,17,86,192,30,189,148,48,269,89,52,31,299,112,83,81,117,88,44,77,186,130,89,334,545,87,44,92,129,372,109,87,101,48,52,82,93,41,222,63,561,34,80,17,327,251,86,204,25,27,235,193,301,85,72,694,29,243,45,77,109,342,73,42,72,17,35,48,114,467,113,42,103,100,82,99,30,102,224,753,111,102,30,510,56,94,99,83,631,83,187,82,33,178,77,172,62,317,63,29,114,53,78,73,137,308,60,130,97,357,97,85,101,121,86,49,70,228,48,718,490,49,139,123,33,101,19,81,176,183,48,262,45,39,108,160,95,51,16,85,68,49,49,26,100,81,85,243,165,28,77,16,21,72,21,51,19,184,383,240,466,58,367,59,539,18,58,41,106,466,350,231,231,65,287,86,67,35,160,96,73,17,49,48,357,401,218,464,60,595,114,66,66,553,69,103,78,173,45,59,124,37,158,308,183,111,79,98,48,141,298,69,224,93,276,80,64,56,190,106,25,192,59,126,23,99,45,94,57,27,86,32,89,119,94,60,77,54,26,287,60,17,99,86,244,101,50,105,17,277,117,26,171,97,206,134,152,82,42,238,338,55,57,149,18,173,89,243,130,29,282,313,170,214,218,119,159,253,89,76,201,289,283,461,346,668,52,69,107,20,178,101,78,100,53,537,117,205,171,103,87,170,36,17,83,152,68,49,192,186,204,97,481,70,67,66,208,83,236,22,74,108,40,94,93,183,105,31,56,52,381,46,150,117,63,54,286,52,83,178,28,162,145,163,51,61,168,95,58,17,75,81,126,230,116,27,17,173,95,63,47,458,46,31,195,104,70,68,28,313,82,173,85,35,101,17,221,268,83,2376,78,86,33,46,21,54,117,342,129,62,76,77,64,357,89,97,116,51,449,218,21,49,28,203,33,45,337,83,813,188,285,616,53,28,104,135,24,206,305,24,267,16,150,131,76,308,22,17,110,142,66,78,110,73,566,72,507,369,186,78,683,49,94,190,28,177,42,150,171,72,306,120,119,76,51,44,79,197,77,437,780,34,87,98,61,248,76,117,190,65,107,119,70,81,41,169,19,73,46,49,51,21,31,91,95,81,285,24,36,78,100,233,65,150,93,85,73,31,47,62,17,48,74,129,118,94,61,54,33,688,80,57,150,90,53,230,186,79,45,225,111,90,504,78,825,91,116,39,224,184,428,87,66,260,205,75,54,19,48,108,99,118,41,80,234,78,468,97,59,90,233,315,19,84,21,175,111,95,91,187,69,104,122,99,83,149,73,87,51,215,123,266,45,73,165,170,120,151,70,95,64,618,80,19,35,186,69,82,48,69,58,189,74,530,88,335,152,146,457,42,269,129,27,68,116,133,148,163,195,176,17,149,48,250,59,18,54,44,211,88,62,314,249,92,43,53,72,101,17,250,49,76,92,102,30,86,36,109,141,62,437,588,19,252,58,64,108,116,257,120,189,285,88,82,70,30,145,197,214,77,119,105],\"type\":\"histogram\"}],                        {\"template\":{\"data\":{\"bar\":[{\"error_x\":{\"color\":\"#2a3f5f\"},\"error_y\":{\"color\":\"#2a3f5f\"},\"marker\":{\"line\":{\"color\":\"#E5ECF6\",\"width\":0.5},\"pattern\":{\"fillmode\":\"overlay\",\"size\":10,\"solidity\":0.2}},\"type\":\"bar\"}],\"barpolar\":[{\"marker\":{\"line\":{\"color\":\"#E5ECF6\",\"width\":0.5},\"pattern\":{\"fillmode\":\"overlay\",\"size\":10,\"solidity\":0.2}},\"type\":\"barpolar\"}],\"carpet\":[{\"aaxis\":{\"endlinecolor\":\"#2a3f5f\",\"gridcolor\":\"white\",\"linecolor\":\"white\",\"minorgridcolor\":\"white\",\"startlinecolor\":\"#2a3f5f\"},\"baxis\":{\"endlinecolor\":\"#2a3f5f\",\"gridcolor\":\"white\",\"linecolor\":\"white\",\"minorgridcolor\":\"white\",\"startlinecolor\":\"#2a3f5f\"},\"type\":\"carpet\"}],\"choropleth\":[{\"colorbar\":{\"outlinewidth\":0,\"ticks\":\"\"},\"type\":\"choropleth\"}],\"contour\":[{\"colorbar\":{\"outlinewidth\":0,\"ticks\":\"\"},\"colorscale\":[[0.0,\"#0d0887\"],[0.1111111111111111,\"#46039f\"],[0.2222222222222222,\"#7201a8\"],[0.3333333333333333,\"#9c179e\"],[0.4444444444444444,\"#bd3786\"],[0.5555555555555556,\"#d8576b\"],[0.6666666666666666,\"#ed7953\"],[0.7777777777777778,\"#fb9f3a\"],[0.8888888888888888,\"#fdca26\"],[1.0,\"#f0f921\"]],\"type\":\"contour\"}],\"contourcarpet\":[{\"colorbar\":{\"outlinewidth\":0,\"ticks\":\"\"},\"type\":\"contourcarpet\"}],\"heatmap\":[{\"colorbar\":{\"outlinewidth\":0,\"ticks\":\"\"},\"colorscale\":[[0.0,\"#0d0887\"],[0.1111111111111111,\"#46039f\"],[0.2222222222222222,\"#7201a8\"],[0.3333333333333333,\"#9c179e\"],[0.4444444444444444,\"#bd3786\"],[0.5555555555555556,\"#d8576b\"],[0.6666666666666666,\"#ed7953\"],[0.7777777777777778,\"#fb9f3a\"],[0.8888888888888888,\"#fdca26\"],[1.0,\"#f0f921\"]],\"type\":\"heatmap\"}],\"heatmapgl\":[{\"colorbar\":{\"outlinewidth\":0,\"ticks\":\"\"},\"colorscale\":[[0.0,\"#0d0887\"],[0.1111111111111111,\"#46039f\"],[0.2222222222222222,\"#7201a8\"],[0.3333333333333333,\"#9c179e\"],[0.4444444444444444,\"#bd3786\"],[0.5555555555555556,\"#d8576b\"],[0.6666666666666666,\"#ed7953\"],[0.7777777777777778,\"#fb9f3a\"],[0.8888888888888888,\"#fdca26\"],[1.0,\"#f0f921\"]],\"type\":\"heatmapgl\"}],\"histogram\":[{\"marker\":{\"pattern\":{\"fillmode\":\"overlay\",\"size\":10,\"solidity\":0.2}},\"type\":\"histogram\"}],\"histogram2d\":[{\"colorbar\":{\"outlinewidth\":0,\"ticks\":\"\"},\"colorscale\":[[0.0,\"#0d0887\"],[0.1111111111111111,\"#46039f\"],[0.2222222222222222,\"#7201a8\"],[0.3333333333333333,\"#9c179e\"],[0.4444444444444444,\"#bd3786\"],[0.5555555555555556,\"#d8576b\"],[0.6666666666666666,\"#ed7953\"],[0.7777777777777778,\"#fb9f3a\"],[0.8888888888888888,\"#fdca26\"],[1.0,\"#f0f921\"]],\"type\":\"histogram2d\"}],\"histogram2dcontour\":[{\"colorbar\":{\"outlinewidth\":0,\"ticks\":\"\"},\"colorscale\":[[0.0,\"#0d0887\"],[0.1111111111111111,\"#46039f\"],[0.2222222222222222,\"#7201a8\"],[0.3333333333333333,\"#9c179e\"],[0.4444444444444444,\"#bd3786\"],[0.5555555555555556,\"#d8576b\"],[0.6666666666666666,\"#ed7953\"],[0.7777777777777778,\"#fb9f3a\"],[0.8888888888888888,\"#fdca26\"],[1.0,\"#f0f921\"]],\"type\":\"histogram2dcontour\"}],\"mesh3d\":[{\"colorbar\":{\"outlinewidth\":0,\"ticks\":\"\"},\"type\":\"mesh3d\"}],\"parcoords\":[{\"line\":{\"colorbar\":{\"outlinewidth\":0,\"ticks\":\"\"}},\"type\":\"parcoords\"}],\"pie\":[{\"automargin\":true,\"type\":\"pie\"}],\"scatter\":[{\"marker\":{\"colorbar\":{\"outlinewidth\":0,\"ticks\":\"\"}},\"type\":\"scatter\"}],\"scatter3d\":[{\"line\":{\"colorbar\":{\"outlinewidth\":0,\"ticks\":\"\"}},\"marker\":{\"colorbar\":{\"outlinewidth\":0,\"ticks\":\"\"}},\"type\":\"scatter3d\"}],\"scattercarpet\":[{\"marker\":{\"colorbar\":{\"outlinewidth\":0,\"ticks\":\"\"}},\"type\":\"scattercarpet\"}],\"scattergeo\":[{\"marker\":{\"colorbar\":{\"outlinewidth\":0,\"ticks\":\"\"}},\"type\":\"scattergeo\"}],\"scattergl\":[{\"marker\":{\"colorbar\":{\"outlinewidth\":0,\"ticks\":\"\"}},\"type\":\"scattergl\"}],\"scattermapbox\":[{\"marker\":{\"colorbar\":{\"outlinewidth\":0,\"ticks\":\"\"}},\"type\":\"scattermapbox\"}],\"scatterpolar\":[{\"marker\":{\"colorbar\":{\"outlinewidth\":0,\"ticks\":\"\"}},\"type\":\"scatterpolar\"}],\"scatterpolargl\":[{\"marker\":{\"colorbar\":{\"outlinewidth\":0,\"ticks\":\"\"}},\"type\":\"scatterpolargl\"}],\"scatterternary\":[{\"marker\":{\"colorbar\":{\"outlinewidth\":0,\"ticks\":\"\"}},\"type\":\"scatterternary\"}],\"surface\":[{\"colorbar\":{\"outlinewidth\":0,\"ticks\":\"\"},\"colorscale\":[[0.0,\"#0d0887\"],[0.1111111111111111,\"#46039f\"],[0.2222222222222222,\"#7201a8\"],[0.3333333333333333,\"#9c179e\"],[0.4444444444444444,\"#bd3786\"],[0.5555555555555556,\"#d8576b\"],[0.6666666666666666,\"#ed7953\"],[0.7777777777777778,\"#fb9f3a\"],[0.8888888888888888,\"#fdca26\"],[1.0,\"#f0f921\"]],\"type\":\"surface\"}],\"table\":[{\"cells\":{\"fill\":{\"color\":\"#EBF0F8\"},\"line\":{\"color\":\"white\"}},\"header\":{\"fill\":{\"color\":\"#C8D4E3\"},\"line\":{\"color\":\"white\"}},\"type\":\"table\"}]},\"layout\":{\"annotationdefaults\":{\"arrowcolor\":\"#2a3f5f\",\"arrowhead\":0,\"arrowwidth\":1},\"autotypenumbers\":\"strict\",\"coloraxis\":{\"colorbar\":{\"outlinewidth\":0,\"ticks\":\"\"}},\"colorscale\":{\"diverging\":[[0,\"#8e0152\"],[0.1,\"#c51b7d\"],[0.2,\"#de77ae\"],[0.3,\"#f1b6da\"],[0.4,\"#fde0ef\"],[0.5,\"#f7f7f7\"],[0.6,\"#e6f5d0\"],[0.7,\"#b8e186\"],[0.8,\"#7fbc41\"],[0.9,\"#4d9221\"],[1,\"#276419\"]],\"sequential\":[[0.0,\"#0d0887\"],[0.1111111111111111,\"#46039f\"],[0.2222222222222222,\"#7201a8\"],[0.3333333333333333,\"#9c179e\"],[0.4444444444444444,\"#bd3786\"],[0.5555555555555556,\"#d8576b\"],[0.6666666666666666,\"#ed7953\"],[0.7777777777777778,\"#fb9f3a\"],[0.8888888888888888,\"#fdca26\"],[1.0,\"#f0f921\"]],\"sequentialminus\":[[0.0,\"#0d0887\"],[0.1111111111111111,\"#46039f\"],[0.2222222222222222,\"#7201a8\"],[0.3333333333333333,\"#9c179e\"],[0.4444444444444444,\"#bd3786\"],[0.5555555555555556,\"#d8576b\"],[0.6666666666666666,\"#ed7953\"],[0.7777777777777778,\"#fb9f3a\"],[0.8888888888888888,\"#fdca26\"],[1.0,\"#f0f921\"]]},\"colorway\":[\"#636efa\",\"#EF553B\",\"#00cc96\",\"#ab63fa\",\"#FFA15A\",\"#19d3f3\",\"#FF6692\",\"#B6E880\",\"#FF97FF\",\"#FECB52\"],\"font\":{\"color\":\"#2a3f5f\"},\"geo\":{\"bgcolor\":\"white\",\"lakecolor\":\"white\",\"landcolor\":\"#E5ECF6\",\"showlakes\":true,\"showland\":true,\"subunitcolor\":\"white\"},\"hoverlabel\":{\"align\":\"left\"},\"hovermode\":\"closest\",\"mapbox\":{\"style\":\"light\"},\"paper_bgcolor\":\"white\",\"plot_bgcolor\":\"#E5ECF6\",\"polar\":{\"angularaxis\":{\"gridcolor\":\"white\",\"linecolor\":\"white\",\"ticks\":\"\"},\"bgcolor\":\"#E5ECF6\",\"radialaxis\":{\"gridcolor\":\"white\",\"linecolor\":\"white\",\"ticks\":\"\"}},\"scene\":{\"xaxis\":{\"backgroundcolor\":\"#E5ECF6\",\"gridcolor\":\"white\",\"gridwidth\":2,\"linecolor\":\"white\",\"showbackground\":true,\"ticks\":\"\",\"zerolinecolor\":\"white\"},\"yaxis\":{\"backgroundcolor\":\"#E5ECF6\",\"gridcolor\":\"white\",\"gridwidth\":2,\"linecolor\":\"white\",\"showbackground\":true,\"ticks\":\"\",\"zerolinecolor\":\"white\"},\"zaxis\":{\"backgroundcolor\":\"#E5ECF6\",\"gridcolor\":\"white\",\"gridwidth\":2,\"linecolor\":\"white\",\"showbackground\":true,\"ticks\":\"\",\"zerolinecolor\":\"white\"}},\"shapedefaults\":{\"line\":{\"color\":\"#2a3f5f\"}},\"ternary\":{\"aaxis\":{\"gridcolor\":\"white\",\"linecolor\":\"white\",\"ticks\":\"\"},\"baxis\":{\"gridcolor\":\"white\",\"linecolor\":\"white\",\"ticks\":\"\"},\"bgcolor\":\"#E5ECF6\",\"caxis\":{\"gridcolor\":\"white\",\"linecolor\":\"white\",\"ticks\":\"\"}},\"title\":{\"x\":0.05},\"xaxis\":{\"automargin\":true,\"gridcolor\":\"white\",\"linecolor\":\"white\",\"ticks\":\"\",\"title\":{\"standoff\":15},\"zerolinecolor\":\"white\",\"zerolinewidth\":2},\"yaxis\":{\"automargin\":true,\"gridcolor\":\"white\",\"linecolor\":\"white\",\"ticks\":\"\",\"title\":{\"standoff\":15},\"zerolinecolor\":\"white\",\"zerolinewidth\":2}}}},                        {\"responsive\": true}                    ).then(function(){\n",
              "                            \n",
              "var gd = document.getElementById('f28cd68b-df07-4f3d-ad76-c8570767d45b');\n",
              "var x = new MutationObserver(function (mutations, observer) {{\n",
              "        var display = window.getComputedStyle(gd).display;\n",
              "        if (!display || display === 'none') {{\n",
              "            console.log([gd, 'removed!']);\n",
              "            Plotly.purge(gd);\n",
              "            observer.disconnect();\n",
              "        }}\n",
              "}});\n",
              "\n",
              "// Listen for the removal of the full notebook cells\n",
              "var notebookContainer = gd.closest('#notebook-container');\n",
              "if (notebookContainer) {{\n",
              "    x.observe(notebookContainer, {childList: true});\n",
              "}}\n",
              "\n",
              "// Listen for the clearing of the current output cell\n",
              "var outputEl = gd.closest('.output');\n",
              "if (outputEl) {{\n",
              "    x.observe(outputEl, {childList: true});\n",
              "}}\n",
              "\n",
              "                        })                };                            </script>        </div>\n",
              "</body>\n",
              "</html>"
            ]
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "trimmed_data_corpus = np.hstack(x_train)\n",
        "print(\"Number of unique words in training data: \", len(np.unique(trimmed_data_corpus)))\n",
        "print(\"Min index in training data: \", min(trimmed_data_corpus))\n",
        "print(\"Max index in training data: \", max(trimmed_data_corpus))"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "CJxcgoPKbRG7",
        "outputId": "f8f10176-ce20-43ef-8dd4-478a475f869f"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Number of unique words in training data:  78\n",
            "Min index in training data:  52.0\n",
            "Max index in training data:  129.0\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "MAXLEN = 512"
      ],
      "metadata": {
        "id": "XI1m3tpEayk3"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "x_train = keras.preprocessing.sequence.pad_sequences(x_train, padding='post', maxlen=MAXLEN)\n",
        "x_test = keras.preprocessing.sequence.pad_sequences(x_test, padding='post', maxlen=MAXLEN)\n",
        "print(x_train.shape)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "ggSxxt_tBwMF",
        "outputId": "bbdb553f-7564-4dbf-8034-010c8b50509f"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "(8982, 512)\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "NUM_CLASSES = len(np.unique(y_train))\n",
        "y_train = tf.one_hot(y_train, NUM_CLASSES)\n",
        "y_test = tf.one_hot(y_test, NUM_CLASSES)\n",
        "print(y_train.shape)\n",
        "print(y_test.shape)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "vF1l-ds1B-mj",
        "outputId": "82e0646c-751a-4f8c-ca13-c93e94e426c0"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "(8982, 46)\n",
            "(2246, 46)\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "#Check: test tensor conversion\n",
        "_ = tf.convert_to_tensor(x_train)\n",
        "_ = tf.convert_to_tensor(y_train)"
      ],
      "metadata": {
        "id": "Wsn4d9aCCXt0"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "Build Model\n",
        "-----------\n",
        "\n",
        "**Metrics**\n",
        " - BinaryCrossEntropy --> only 0 or 1 labels\n",
        " - CategoricalCrossentropy --> multi label classification, tf.one_hot y labels\n",
        " - SparseCategoricalCrossentropy --> when passing labels as integers"
      ],
      "metadata": {
        "id": "G0abn44cZquU"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "keras.backend.clear_session()\n",
        "\n",
        "model = keras.Sequential()\n",
        "# model.add(layers.InputLayer(input_shape=(MAXLEN)))\n",
        "# model.add(layers.Dense(512, activation=\"relu\"))\n",
        "# model.add(layers.Dropout(0.5))\n",
        "# model.add(layers.Dense(NUM_CLASSES, activation=\"softmax\"))\n",
        "\n",
        "model.add(layers.InputLayer(input_shape=(MAXLEN)))\n",
        "model.add(layers.Embedding(MAXLEN + 1, MAXLEN))\n",
        "model.add(layers.Dropout(0.2))\n",
        "model.add(layers.GlobalAveragePooling1D())\n",
        "model.add(layers.Dropout(0.2))\n",
        "model.add(layers.Dense(NUM_CLASSES, activation=\"softmax\"))\n",
        "\n",
        "model.compile(\n",
        "    loss=\"categorical_crossentropy\",\n",
        "    optimizer=\"adam\",\n",
        "    metrics=[\"categorical_accuracy\"]\n",
        ")\n",
        "\n",
        "model.summary()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "jNl5c7hzDLtT",
        "outputId": "830c8434-2f08-4272-f98e-99da0868e24d"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Model: \"sequential\"\n",
            "_________________________________________________________________\n",
            " Layer (type)                Output Shape              Param #   \n",
            "=================================================================\n",
            " embedding (Embedding)       (None, 512, 512)          262656    \n",
            "                                                                 \n",
            " dropout (Dropout)           (None, 512, 512)          0         \n",
            "                                                                 \n",
            " global_average_pooling1d (G  (None, 512)              0         \n",
            " lobalAveragePooling1D)                                          \n",
            "                                                                 \n",
            " dropout_1 (Dropout)         (None, 512)               0         \n",
            "                                                                 \n",
            " dense (Dense)               (None, 46)                23598     \n",
            "                                                                 \n",
            "=================================================================\n",
            "Total params: 286,254\n",
            "Trainable params: 286,254\n",
            "Non-trainable params: 0\n",
            "_________________________________________________________________\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "x_train.shape, x_test.shape, y_train.shape, y_test.shape"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "yyEh6KvrzZDA",
        "outputId": "cb0fa39e-d784-4781-8941-fa07a6dc1388"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "((8982, 512), (2246, 512), TensorShape([8982, 46]), TensorShape([2246, 46]))"
            ]
          },
          "metadata": {},
          "execution_count": 110
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "%%time\n",
        "history = model.fit(\n",
        "    x=x_train,\n",
        "    y=y_train,\n",
        "    epochs=10,\n",
        "    validation_data=(x_test, y_test),\n",
        "    verbose=1\n",
        ")"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "vBJH4xaREGcx",
        "outputId": "a53ac3f7-9f6f-4389-96c7-f6e2238137c7"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Epoch 1/10\n",
            "281/281 [==============================] - 38s 133ms/step - loss: 2.4376 - categorical_accuracy: 0.3725 - val_loss: 2.1518 - val_categorical_accuracy: 0.4150\n",
            "Epoch 2/10\n",
            "281/281 [==============================] - 38s 135ms/step - loss: 2.0447 - categorical_accuracy: 0.4685 - val_loss: 1.9388 - val_categorical_accuracy: 0.5365\n",
            "Epoch 3/10\n",
            "281/281 [==============================] - 37s 132ms/step - loss: 1.8337 - categorical_accuracy: 0.5616 - val_loss: 1.7662 - val_categorical_accuracy: 0.5980\n",
            "Epoch 4/10\n",
            "281/281 [==============================] - 42s 151ms/step - loss: 1.6634 - categorical_accuracy: 0.6146 - val_loss: 1.6375 - val_categorical_accuracy: 0.6300\n",
            "Epoch 5/10\n",
            "281/281 [==============================] - 40s 143ms/step - loss: 1.5373 - categorical_accuracy: 0.6480 - val_loss: 1.5402 - val_categorical_accuracy: 0.6594\n",
            "Epoch 6/10\n",
            "281/281 [==============================] - 37s 132ms/step - loss: 1.4377 - categorical_accuracy: 0.6696 - val_loss: 1.4735 - val_categorical_accuracy: 0.6741\n",
            "Epoch 7/10\n",
            "281/281 [==============================] - 38s 134ms/step - loss: 1.3677 - categorical_accuracy: 0.6867 - val_loss: 1.4166 - val_categorical_accuracy: 0.6803\n",
            "Epoch 8/10\n",
            "281/281 [==============================] - 37s 132ms/step - loss: 1.3082 - categorical_accuracy: 0.7020 - val_loss: 1.3756 - val_categorical_accuracy: 0.6901\n",
            "Epoch 9/10\n",
            "281/281 [==============================] - 37s 132ms/step - loss: 1.2575 - categorical_accuracy: 0.7136 - val_loss: 1.3475 - val_categorical_accuracy: 0.6915\n",
            "Epoch 10/10\n",
            "281/281 [==============================] - 37s 132ms/step - loss: 1.2175 - categorical_accuracy: 0.7247 - val_loss: 1.3111 - val_categorical_accuracy: 0.7150\n",
            "CPU times: user 10min 21s, sys: 30.9 s, total: 10min 52s\n",
            "Wall time: 6min 22s\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "score = model.evaluate(x_test, y_test, verbose=0)\n",
        "print(\"Validation loss: \", score[0])\n",
        "print(\"Categorical accuracy: \", score[1])"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "oHR_b9j3FO7d",
        "outputId": "9327c234-bccf-4f4d-cc1e-720f84f6b029"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Validation loss:  1.3111015558242798\n",
            "Categorical accuracy:  0.7150489687919617\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "y_prob = model.predict(x_test)\n",
        "y_pred = np.asarray([np.argmax(i) for i in y_prob])\n",
        "print(\"Labels in y_test: \", np.unique(y_pred))\n",
        "y_test_ = np.asarray([np.argmax(i) for i in y_test])\n",
        "print(\"Labels in y_pred: \", np.unique(y_test_))"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "Ro0HU2ccFZx0",
        "outputId": "e6dd5d58-0f6e-45f9-d5d8-33b5cdc8154b"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Labels in y_test:  [ 0  1  3  4  8  9 10 11 13 16 18 19 20 21 25 30 36]\n",
            "Labels in y_pred:  [ 0  1  2  3  4  5  6  7  8  9 10 11 12 13 14 15 16 17 18 19 20 21 22 23\n",
            " 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45]\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "cm = confusion_matrix(y_test_, y_pred)\n",
        "disp = ConfusionMatrixDisplay(cm, )\n",
        "disp.plot()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 300
        },
        "id": "d88eqTLt00DO",
        "outputId": "f7836804-a434-47b2-ea10-b44d85143eed"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<sklearn.metrics._plot.confusion_matrix.ConfusionMatrixDisplay at 0x7f9152168c90>"
            ]
          },
          "metadata": {},
          "execution_count": 126
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAT8AAAEJCAYAAADihSAbAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4yLjIsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+WH4yJAAAgAElEQVR4nOydd3hU1fa/332mZSYz6T0hhBCSQIgIhKaogDTFrljBcq0027Vc69de8f4Uu16vXRS7V1FRFCsqKlWE0BJCCul9JtP274+TxJRJMkkmJCZZPPNAdvbeZ63zJId91vrsdwspJYM2aIM2aAPNlN52YNAGbdAGrTds8OE3aIM2aAPSBh9+gzZogzYgbfDhN2iDNmgD0gYffoM2aIM2IG3w4TdogzZoA9K0ve2At+YuSJ4LPAZogP8ArwOvAJGABJ4Ddjbto0RlPtBynBKV+UD9fBrgVyAXeMKbcV60vQ1MbuqTEpX5mLsg+b/ACUChEpU52lM8nZirQ7+ATT6Kp625uhJPl/06BPEc0rn6gg/t/W4MFBO9ofMTQjS76VLKdm96/YMqE5gFHAA2AMuAaiUq83d3QbIF+A3wA6Y16XMe8GGLcecoUZnb3QXJ1wIZQAAwskUfT+O8adsI3K5EZb7bxKdTgDCgGnhFicoc3UY83sx1mpd+BXtxH7xtaznXOV2Mpzt+9WQ8vTFXX/Chzd8NBogd8oefEMLTL8o5Uso2b/pRGRZ5+3WhnH5OJQDXLDMC8P8etzb2Wf1+AEajYPrcisY+cbEKCUM1rca99U4dTz9m5pHHrNx8g4maWtmsj6dx3rY19euNFy0896KNdd86iI9TePPVAI6YXs6E8Vr+dZ2p03N9872D2TP17Y575AF/jjlKR8aR5d2Ox9NcDT51Np6u+tXT8RzqufqCD57ali3TcPfNwTcrUZn300WbM91flpS6vOr725a6z6WUc7t6LV9Yb7z2TgR2Syn3Aggh3gROBtp8+A2N8seeF8UkMQEAY34pqeNqmSTiAIiMs5M2IpNfvgpgksho7HPY5DKK8/QtxlmZcbeDN++NJMXsItQvj7LdxmZ9PI3ztq3Br8g4Oxnpu3l+43gmCQ2Rwo6JfUwS45kaXY49r6rTc237JAp7Xk2740KdB9A5q5kkxnc7Hk9zNfrUyXi66ldPx3Oo5+oLPnhqK8z/HiCWblhJqYtfPo/3qq8meldYd67lC+uNgkcskNPk6wN046b7mVzc9p8s1rwVjNMhOuwfGWenvFjL7q2mrl7Sa5+euT2G2mqNz+aqsw3Wpwat75oE3F7+6QvWG6+9ZwBzpZSX1H+9EJgkpVzaot9lwGUAR48PGv/sfUMIj3FgDnDjtAvys/Vcc0oSK1ZnEhTqIjDExZ4/DOj9QKNI8rL05O/XEze8jupyLQmpNixBTorydEQn2DEY3QgBQkBNpUJFqa7ZuJSxtRj9ZafbSgu1HH1iBXU2hXeeCWfVE5FkTKtk6f25hEU7eGV5JFvXm1nwzwJuW5DI459lojdINv/o32yuonwtEbEOgsKcWGs0vPV4BDs3mVh874F2fago1WAJcVJTrkWnlwRHOCnM1SIUOvQ9Is5O9FAHuXsNXD4jhaX35zB5diW2Wk1jny3rzezZZmwVjzd+hcU4ms3lzT3t6ri+Oldf8MFT2/asamZO13frtXfcGIP84bMYr/qaYrJ+k1JmdPVavrDeWErkAnOFEIVCiG1AXH1bM5NSPielzJBSZmzbIkgcacPtEpw7NhVFK6mtUbjv9b3s2Wpi2dxkDuZoGZpSx6PXxbJo1gjSp9Tw+zcWUg63krnZyBXHJlNaqOXVR6KoLtdyWuph3HT2cMqLNfgHuJuN2/itmcSRti60VXPMSRV881EQ544dxfSTyxmaYmXJfbksv2oIuXv1TD+5HGuNQuwwOwuvK+DAHgOhUQ5WvxbabK7UcbVUlmr47I1Q/nFkKideWIzNKjr0ITHNhsnfzfJrhrDs+BGUFWrQGyTDR3Ucz9p3gykv0aDVudHq3KRlVBMS7mzWZ3+moVU83vllJSTC2el72tVxfXWuvuCDp7Yjp+gAtnX3l3tw5dfeBYXQAvtRV3UPAk7gXCnlH22NOTYjQj77UBTxyTaK8nTk7dMzJMVNRJSVfXsCcLsFQxMqKMrX4XIIFA3kZ+nZ/ruZExYWU1OloCiw5q1Q1rwdymMf7WTR7JEMT6vllmf2UVWmQUoax+Vl60kda8Xo7+pUm6MOhqbY2bvdDykhKMzJ7q1GRqRbkUBgiBObVWHjdxbWfx7IlQ/mUF2hwVqt4ZFrhrDk3tzGuarKNKSMtTbOFTXEztcfBJFyeMd+lRTqGJVRi6JI1r4bzLRTykCqq9z2xm3+0UxpoZarHsqlKE/Hxm/NJHu43pFzK5vFs3urH/4W2a5flaUawqId2GqVTt3Tro7rq3P1BR88te3IqmbGtO6t/MaO0ctvPo3yqm9gbE6vr/wOecFDSukUQlyCqq2LBe5q78EHEB2lsHOTic9XhnDhvwoIDLZSWmbi+GNObezz+rsfsX2DPw8tGwrAsaeXMnFWNYW5erIz/UgcZSUy3k51hYZ3no3k1Z+3UWdTyM7048AeA49eN6Rx3PRTy9i9zdjptkV3HwBRzaKZKc36/Lw2oNm41HFWpsyp4J+njsBkdnHGFUWERjlazZ86rpYnb1ELHsvf3832Df5otHToV+o4K5cek8oTn+3kjCuK2PidmYpSrRfjaln3QXDja+/UeeUoHq7XMp7pp5axO1vf4b0Za67msmmpnb6nXRnXV+fqCz54aks4/XtmTNN3q+AB4Obvg8jrlQy6lHI1MAPYJaW811MfIcRlQohfhRC/unCg1bmZMqeSCyaN5KlbY9Bq3Uyftb/d6wghSRpdy8evhrNk7khstQoLr81nyuxyLpiSxrnj09HrJTEJdT0QZdvWmaJLVwsebrdg8exUzssYRcwwO+ZAZ3dcHrRB69Ak4EJ69ekL1isiZwAhRALwsZRydEd9j82IkC8/GU5gsIuKEi15WXqsBDB+UiGFBSZcLvW1NzuzecFj11Z/5pxTQk2lBo0GNn5nIfnwGrIz/TD4SRJSrAQEO1E0kvKS7hc8Kko1xCfX4ahTKC/W8tV7QYTH2luN0xkkMcPsuJySsCgnQoHfvzUTGOpst3jiTWEhL0vPlp8svP1UBI9/mklJgY6Sg1omTK/EZm0/Mb5lvZmSAh1XP3yAojwdv31rIWVsTZcKP/29QNDf4vFFwePwMXr5xafhXvWNiM3r9dfeQ77yE0IMEUJ8DXwBJAkhrupozKYtTiJjHVSWa7j6xOGkT6nBrKvAWSe5/qR4Fk8bRnGBplXBY+fvfpgtLp64KYZLj0lmypxySgp0TJ5ZwaYfLFwybRS/rjNjsvim4JGYZsPlFDxybRwIybSTy1sVMtKn1PDec+EsyEjjw/9GsPUnM5WlWu64cFiHxRNvCgvpU2ooytVy5pKD5OwyoCiS+CQrIZEdJ8b3Zxo4/4YCDh7Qcem0FNImVHtVKBmIBYL+Fo8vCh4ScEjp1acvWG8UPKKBaKAUWI2adzylvR0ex2ZEyKfuiyYyrg6jxY2zTiE3S48lyMUnr4Zy4oUlhIQ7kBLqrAoVpdrGpO7wNBtanUSrk2i0kp++CGTeghLKijS4nApF+TqCwxz4mdzdThrnZ+kpOagjfXI1YdFOXv93JJt+MDcrZORn6dm83sLad4O57tH9/PylhbOWFHHnPxKa9WureBIS4ezQh8J8HcecWEF5sYagMBfrPgxiRHrHvkfG24mIdaAokrIiHbn7dMQl2gcLBAMgHl8UPA4bo5OfrPZOuxwfV9DrK7/eKHjkCyH+jbofMQy12nsx8M+2xkRHKezeZmTp3GTgr6T+xGMrOeqECsqLtPz8hcVjMSA7069Z28RZ1eTuMzQWQbQ6yd7tRh65Jr7ZuK4kjRuKBjfOT+KuV/axckUkU+eVeyxIXHFnLv+5JwaT2UXmZlOrgkdbxRNvfFjwz4Ncf0ZSYzHlt28CEKLjQsn0U8vY+pO5mQ/B4a7BAsEAiMcnBQ8Jrr6xqPPKei3nB415v2+B0VLKyhbfaxQ5nz0vZPzV5w0lKt6BRpHs2mqkokTHlDkVVFdoOJij46UHo7HbVJlLUZ6OqfMqcDnhl7UB1FRqURTJf77bgZ/JTVCok1/XBRAzTC10VBRrCApzoiiwa6sRk9kFgmbX89Q2KqOGgBAX+VlqdfTY00s56sRyElLqGgXAefsMzD23pNm4keNrCQxxkp9t4MlbYznjiiLWvBXcrF9ZsZawaAf2OqVdH1q2lRzUYglyUWfVYAl2EhTm5NevLRjNrg7jMZldFB/UMSLdSkm+jvz9emaeUYbToeYwVz0ZzvRTy7y6Ny3buhpPV8f11bn6gg+e2n7aWsVpJxmeVKIyl9JFSz9MJz/0cuU3fMgAXPkJIfxQH3hGIAn4qOWDD1SRMyqmilkTI2T65Boun55Ccb6OV3/5k3UfBuFywQ3zk6gs03L18ixmnVHOP6amYqtVOOnCYt56MoKjTijnszfCOOWSYmy1aoqzoFbP/14KZcPXgZxwfhFL7s3l1eWRrHoygld/+ZPN601MmVPV7Hqe2rIyDfjluTGZ1f9AwmPspE2o5c6LElj2wAGmn1zOqqfCael7VYWCTq8hdriNm57KxmRxoWhcpE+ubez3xsY/8A9wc/HU1HZ9aNn2zvY/kNJNVbkWnR4UjZtxx1SiN8Clx7Qfz+b1/oyeUMPurSZMZheaejH5nRcN4/rHcgiLdlBwQMes+eUd3puWbV2Np6vj+upcfcEHT20v/LIJwNHN325cdLzF1KuZhEgB3mrSlAjcjoqxewtIALKAM6WUZUIIgUqKOh6oBS6UUv7e3jV6Q+pSB8wG8oDbgAQhxOSOBgmg6SJVq5Mo9ffZYHSRengt0i2oswkQEB7jYFiqjbjhdXyet4nL78hleJqVkhIjP64J4Y9NQWhCQ0iZ6AIJweF/SUEURbS6nqe27J1GwqKcjTsiZp9VRs4uA0V5ekCw7sMgRk+qaTXut3UBXDVvBLl7/Lh/0VA2f2/mjUejmvVr+BHqyIeWbdZqhc9WhnH+pNHcv3goW9dbKC/SeTWXrVYhItbBT18GIBRIm1CDQGAwuhoLONs3mL26Ny3buhpPV8f11bn6gg9ttXXX1IKH8OrT4VxS7pRSHi6lPBwYj/pAex/4F7BWSjkCWFv/NcBxwIj6z2XA0x1dozcKHgJ4GbXgcTPwPbBISvlzW2MWnBAlr1wQQ9QQO4oGdm81Yq9TOOakctwuQEDWn35UVWhY/VooCak2ElJsZO/044tVIVy9PIc/Npg4/fIitv0WwmM3D2VYYjk1lRqS0mu44PoCaqsVnA6F3VuNGP1dCIVm1/PUVl6sJXOzsXFHxM6NJlLG1mIwugkMcVJbraGkQEvJQV2rce88HcFdr+zjyVtiOeOKQtasCuG480oa+5UXawmLcmCvE+364NkvHU/fHssL3/5J5BA7P38ZgN7P3eG4pHQrH7wQzvxFhViCnLz5eCS7txpZcu+BxgJOzh5DMz+99aur8XTnPvTFufqCD57aftlaxSkndu+1N+0wvXzzkwiv+h4Wn+v1a68QYjbwf1LKI4UQO4Fp9bWDaGCdlDJFCPFs/b9X1o9p7NfWvL2x8jsSWAhcDlSh7vII7WhQUZ6ei48ayUVHjOSnNQHUVmm4YMpITkoaw8KJowgKc1JnFfywOojX/x3FT2sCOOGCYl5cv52xR1XhdAj+95J6mZKDBjZ8HciOPbEcMa+W4nwdCyekNc7t6XpttW350ULuXgMXHTGS71cHsvlHM+eOTWPe0DE8c3sMxfk6j+MabMt6M7dfkNhq/szNRqy1ilc+eGpzuwUXTR3F/PTRxAyzY61pf66iXB11VoUP/xvOQ1fGs/lHCytXRLHhqwBunJ9E7l4DK1dEdsoHX8TT3fvQ1+bqCz54avt0jb2jX0GvzC2FVx8grGETQ/3nsnamPRtYWf/vyCYPtAJU0jl0gRbV2wWPINSl7DIp5bYW32tFdWkq0tzzh4nUcbUEhTlACmqqBZWlGkZPqsXo78ZuVdAbXSybm0xotIO7Xs7C7QJFgcf+Fcenr4Wy8NZKxkwsRafU4R+ozv3N/4JwuyUTZlR1LCbuoih4zzYj6VNqSEq3kp+lZ/Vroa2ILd0RsGZlGhk7tarxXsYn26irVSgrblvILRRJQLALP5O6QUlRoLpCw4NL4lsRXG59LgutjnaF3P1dFNzf4vGFyHnUYXr52sfe7e0dP9S7vb1CCD1qiixNSnlQCFEupQxq8v0yKWWwEOJj4AEp5ff17WuBG6WUv7Y1d68B4uqJzl8DMUAroqsnqktTkWZ+to6XH4risukjueHMRIanWRk5zsrzd0Zz3WmJGExuVj4WyZARakX3/efDWLkigvVfBrJjkwWNVjLuiBLuWJZObKKDey4dyqJZIzju3BJ++dLilVC0q6Lg/bsMrHwsgrx9Bq46YYRHYkt3BKx7tvpx/RlJLJ6dyrWnjAAJRnP7Qu6QSCcPXRlPUb6Ouy9NYMNXAZx12GiPRBoh6FDI3d9Fwf0tHt+InAUuFK8+nbDjgN+llAfrvz5Y/7rboBkurG/PBYY0GeeRFtXUeiPnF45aVfoHMAmYCVwgpfy4rTENIuemIs29242MP0Zd3Wi0Er2fm+oKlc4SGOrCbhPcfO5wCg/o1UkUhfQja5BOFxsvSsNy9V6eWrcL6YaaEoXKMg2xiXby9qn/K3ojFO2qKDgoXJWgBIY4KSvSUVGq4Yu3Qpg5v8wnAtb9e4ykT6pG0UhM/i70Rjcl+foOx21eb+H4BcU8fXssx59XysoVkdz+/L4WRBozn68MbZYH9CTk7u+i4P4Wjy9EziMPM8iX/ucdz29ygnc8v3rS++dSyhfrv34YKJFSPiCE+BcQIqW8QQgxD1iKWu2dBKyQUk5sb+7ewNhHA28AQ4FioLi9Bx/8JXJuKSZeNEsVAEfG2Xn800y2/WJm+VXxRMbZiU+2cswpFbz9bAy4XAidjuJcHRPOrON9v1CGu3I4ase5HBeRxQl/bu6SoLmrouAGWkuD78vf301VucZnAtbU8TaWzB0JwLXLs5FSIqV3NBikYMG1B0HCzPklHog0tY15wPaE3P1dFNzf4vGFyFkisEtNd6ZoZkIIf9Szfi5v0vwAsEoIcTGQDZxZ374a9cG3G7UyfFFH8/fGDo8tQogdwAWABbiuO/M1xdibg1RIYp1VEBDiIiLuL/mKlBLcbspmhqsZhD5gTYktogcSEFqdm8mzy3nhvlhGjqv2asy1pyZRUqAnMNTB459kkrXDz/eODVq/NBVj77sfZCllDS2KoVLKEuBYD30lsKQz8/fGa+8JwPFSysVCiGnAdVLKEzz0a7fgsWW9mXkLS7AEu6itVqit0mCzCoz+Eq1Wsu6DIFY9FcGo8TVccXceGkXy9YdBvPTiEMYdUcXS67JRdPC1SGBiWfYhJZdsWW9m359Gbnk2q01iS3eS2SWFekZl1OBvdlFbo7D86gQW37O/4wLOTxb2bDNyxV25aBRJVYWGkAhHMxrMIMa+f8bji4JHymF+8umPhnrV99hhmb2+w6O3pC6XCSHsqGSX44QQr7Xs1FHBI3ungYAQF1+/F8R549JYNDO5sc9t5ydw1rJCDj/aypL787h14XCeuXsop11ajOaZISx+JJ+bLxrOjF1nMT7o4CEnl+zP1HPT09ntElu6nsyuZsL0Sm5dmMTWn83o/SQ2m3c0mPxsLcvuP8CtCxJZevwIhia3psEMYuz7Zzw+KXhIgUsqXn36gvXWoeVZqAeGj6aNlV9T81TwKMjRM29hKVk7DbicAoPRjSJU+onB6EYoYLMKcvcYWDo3hX/+v/2YA1yUFGqJHmrnlnOHA7D0/hymzK48pAlom1Vw5HFV7RJbupNQNwe5WDwrmdc2/MmHL4YRFObwKp7yEi1T5lRSkK1Ho5VYaxSPPgxi7PtfPL4oeCSnG+XjHw3zqu/cxD97feXXGwWPTpvngoeVgv16HHYFJGz63uyR6lJcoFZ7H7kmvrFN3X6mmsspqKnSHPIE9KdvhPRcQv2oauqsGuaPHq3eq7E1XsfzzUdBHfowiLHvf/H4ruDxt3ikAL338JPAmvq/n/XUoQXVxeMkg8n5QRu0vmO+Lnj0tPXWa2+slDJXCBGBmvdbJqX8tq3+x2ZEyCfvi+p0cr6tRO+YI6pxuxSEUMdFDrH3aAJ6yuzKZjsiZs4vbXY+7llLD7baJdGdhHpXz+3ti8n5vuDDQIjHFwWPpHSTfOiDFK/6np60qddfe3vrMV0jhHgHFW0VA8xvr/OmLc7WyfksLcvuz+HW84ax9Lgkj8l5T4nenD0mooc6uHdxIstOHEniKGurs2l9mYD+/RtLqx0Rq56MaHY+rqddEl1PqNef23ttPMvmJVNW5P25vX0xOd8XfBgI8fThHR49Zr0hdfFH5fR9jbpZ+UtguZTy3bbGeCp4eJucb5norSzXERDsJHKIHUUjqSrTEhrZc6j2zT+a2bu9ORll5YpIZp1Z0kiDWfNmSKtdEt1JZnf13N6+mJzvCz4MhHh8UfAYnu4v73t/lFd9zx7xa6+v/Hoj55cEnIFa6b0GeKO9Bx94Lnh4m5xvOS5jRjWRQ+xcc0oqdpvCk59up6ZSw2XTeyYBnTqulreeiGy2IwL+osFcPkN9TWi5S6JbOzy6eG5vX0zO9wUfBkI8vil40GdWdd5Yb6z8Dkdd+W0HxgC/AVfVq7mb9msXY+8N0tsTZj51vA2nU3DCwiKkVJH3Op3Ebu88Crzl/IvuPsCUOZXNfKgo0fLzFwHNRMGrnohk5hmljTSYT1eGtMLddwdRrveTxA2vQ6NI1r4XzFEnVFBdoRAY6up2PIPY9/4Zjy8w9sPSzfKO9w7zqu+FyesH5MpvODABCELdg7cQlcYwp2mnjjD23iC9W2Lmw6IdlBZqmX1WCZcdm0Zxvo6Vv23G6O/m4qM6jwJvOb9G5yY00tFsrnUfBTaKghvQ9j9/EdBIg7l8RgqPr97VCnffdUT5dvRGyZI56tePr85kf6aeI+ZWdhijN/EMYt/7Zzy+wNhLSZ8RMHtjveHpD0C2lDIZFU9dDeg6GtQVpHdLzPy0k8spyNJReEBH/l4NQSG2xoFdwX63nH/c1GrcLtGsT0i4k7wsfTO0/YkXFlN4QI/TodKjPeHuu4oo1+pUlH1VmQanXfDdJ4GkTaxBuoVP4hnEvvfPeHxjAreXn75gvSV1+Q64BJXs8h/gLSnl9W3194Sx9wbp3RIzv+bNEHL2GFhyby5V5RpcDkFetp7kw6xdxMW3xthbgp2txsUm2puh7RVFotWre5DLinT8us5MWJTDZ5j04DAHNdUaFAUsQSreK3+/3mfxDGLf+188vsDYDx1tkTe/O86rvlekftvrr729tUZdBrwOvAmUAfd1NKCrSO+mmPmGYsPPXwZw2bRUFs1Sl/zdQYG3xNh76tMSbf/nbya+fj+YeUPHsCBjFFvXm72Ox5v7YLcrLJmdwqKZKbx4f1TjytNX8Qxi3/tfPL7C2A9KXby5cAs8tYfvt0t1iYizNxMKL70/h8mzK5sJPresN6M3SM5cWkh+lp5PXw9lx0YTD769h7x9BtxuCI104HAIr4WiMQl2GjD5K1dEsO6DYB7/NJPKMi0ajSQ0SpUelBXpvEa8+1pEGxTmpPCACk0NCHaRu0+P3s/doQ9ZO/2YMKOK6Hg7OXsMxA6rw+kUOO1Kp5D1A0EU3N/i8YXIOX50gLzhHe8Wc8tGfj0wV35CiGtQoYMG4NH6s3ybWUdUl7XvBjcTCqdlVLcSK1urFaadUk5+lp4rZqYwaVYlleUaNFrJw1fFceXxIyg9qOmUyPnlByMbMfnZO1W3rTUKl01P4eKjU3G5YPN6/04h3n0rorXh5+/mnWfDWTwrmfJiDeExdq9EzlvXm7ltQSJ5WQaWzkmmtlqDwyY6jawfCKLg/haPb0TO4JBarz59wXpD6hKLelzlr6j01eOA1VLKl9oa40nkvPlHM6WF2sb81MZvzSQf3lzIGTe8jqAwJxqtpKxYx64tJrbXH2FprVGf+xu/M5M8pnOiUJ1BPewnP9tAUJiTgGAnZcVaXvt/sUTE1HLaJUVo9XiNePe1gLXkoI5RGTUoGljzZgjjp1USEuFEyvZFzk3x+tWVGmoqNTx9W2ynkfUDQRTc3+Lxhch5yOhAedXbHR7BDcD1o9b0+sqvtx7BOuBYVDz16XTAVm4LY7/ug+DG196p88pRWlBdMmZUkzymtlHQ/MDKTMJjHNRZFWqqNCDV/606KwqNjLPzyAe72LbBwlO3N4c33vGfnbz0UDTHnVvqNeLd1wLW1HG1XHyUirGPjLNz6qVF/PJVAMuvim9/3HgbT96qfn3t8mx2bzN2CVk/EETB/S0eX4mc3X8jqUtvYOxz6w8huRfYCayRUq7piWtVV2h5++ko7nt9F7ZahT3bjbjdPUeDOXtpHi4nrP8sgOPOLfXJnN0xT4h/b6wBf//fB7w7jGbQBq3BXH1ExuKN9cZrbzDwLnAWUA68DbwjpXytRb8OMfYnXlhCYIiTnD0GdHrZiLFv6LP3T39SDq8hKNwJEkoKdGxe78+4o6qbnfcbHu0dDaahTa93Ywl28daT0ez+w5+bn9iD0ezC7YLTUtMJCnVyz2t7KS7QMSSpDpPFRXmxFgHkZekpK9YwaWY1ZUVakJC/X09Rnq5VMjs+uQ5HXeeLDSUHdaRNqCEsxsHmH8ysfCyyQ8x8XpaePzZYmDy7gsBQ9dX33eci+HND55H1A6FA0N/i8UXBIzYtSC5eNdWrvreO/qTXX3t7Y406EzCigg02oYqcj2jZqWOMvbpaK8jRs3hWSjOMfSNyfZcfKx+P4vJj07jnikTSJ1ezP9PQ6rxfb2gwTdtuWTAMP5Ob/Zl+LLk7m/+9Gs7BAzrcbkFknCoZCAx1svF7MxdOScVocrNzo7Fxri0/mqmzCW5bOIx/njqccUdXeyxcuJxdKTZUM2FGFdk7/fjsjZD6B7t3GPu9f/pxw5YMnL4AACAASURBVJkj2PG7P8/8XxwZ0yq45pGcTiPrB0KBoL/F47uCh8arjzcmhAgSQrwjhNghhPhTCDFFCBEihPhCCLGr/u/g+r5CCLFCCLFbCLFFCNGh4LA3Vn7nAC8CsUAFsB94Xkr5f22N8VTwCAp3kpBqQ7qhrFjHl28HkTGtulVRZOKxVViCnbgcAmutwievhnLG5UWAet6vweTG5RAdFgMa2mKH1TXD5Bce0KPXq/MEhjpxOdW5DH6S0kIta94KZtKsKoYMt1GUp29GernizlwUDTgdgk9eCW12bm9D4SJ9cnWnz8cNjXIQHuNk73Y/AkOdaHWSmgoNbnf7MW5eb+GDF8J4bcN2Lp2eyvJ3dxEQ7MLpEJ1C1g+EAkF/i8cXBY+YtGB58ZvTvOp7z2EfdLjyE0K8DHwnpfxPvTTOBNwMlDY5tzdYSnmjEOJ4VP1ww7m9j0kpJ7U3f28UPJyoUIMf6v9dCNjaG9AWxt7f4qaqQi1cWIJcHhP//zw1CfjrfNz1nwWy9p2QxrbHP830qhjQ0LZ0bnKztuICfbMiyPo1wTx1ezzvbN3EeRPGcOScYoLCnETE2LnoiJHNSC8bvgpo89zehn5dOh/3qGrmxKR59LO9GFPHWXHUKRTl6Xjphz/bpsF0gKwfCAWC/haP7woevsn5CSECgaOBCwGklHbALoQ4GZhW3+1lYB1wI3Ay8Er9EZY/1a8ao6WU+W1eoxdWfiOBD4EpgBVYC/wqpVzWol8rqsvI8bUY/d3YahTWvhvCm09EMCqjlguuzydueB3ZOw1odDSSK4al2ggIcVFRqsFuU1j5WAQ/fBqEf4CL6x7dT8a0KmxWhdy9esxB7m6RMfR6N/6BLrb+bGFIoo2YhDpeejiW3D1aTr+iiJTDazmw28DubUZqKjX8/EUAi+/JJTzWwXf/C+THzwObUV0a/A8Kc+Ln7+bEYYcxdV55qz6eSB9Dk204HWqucNWT4Uw/tcyreBruV3mxlmtPTeLfH+z2igYzECko/S0eX1BdotOC5QUrj/Wq74Nj3s0Gips0PVcPM2n4/fdIfwJypZRB9X0EUCalDBJCfAw8IKX8vv57a4EbpZS/tuVDj+X8hBD/FUIUCiG2NWkLAVYAgcA+VJDpJsDVcnzTnF9xoYb0yTWsuCGOa05KwmBy43RASYGerB1+3PGPYZQWaohLquPW84Zx6bQU0ifVsG+HH7edn0BErIOv3gvmh0+DAFh8Ty4xw+p48YFo7rpkKImjbM3GVVcqpE+u6VTbv84ajp/JTfqkKm69YAR5WXqOPa2EqPg6ho20kp+tCq1Hjqulrg6W3p9LSaGWlx+KYthIGzq9u9X8+3b48fBVQxrvSWmhtkO/4obXUVcnuOU89RStsGgHBQd0XsWzb4cft5yXCEBNpZbtvxlJHVvb7Xuj+mUjNMpxyMb11bn6gg+e2o6YpIXuUl0QuKV3H6C44fe7/vNci+m0wDjgaSnlWKAG+FeLZ4REXXB2yXqy4PESMLdF27+AtVLKSNT9vN+h7u3N7GgyAezYaKKqXE2Wag0CUwAc2OdP8UE/TGY3UjanVJQU6DhrSSGOOsF7z4UDYAqUHHlcBRu/s/Def6JwudU3/26TMYT6jZoKDfl7NaxfE0hhrpakw+rQaGDz+gA0eg1+/m7CIp3oDW72bDXxztMRHqkuDf7XVDZPDnfkl8uh5hsNRldjoWT7BrNX8VSXa3A61X/rDS5SxtR6RYMZiBSU/haPr8yN4tXHCzsAHJBS/lz/9TuoD8ODQohogPq/C+u/n4uKxmuwuPq2Nq1HX3uFEAnAx1LK0fVf70R9X3cBGtSdHg5gspSyvK15mlJddAaJOcDF+i+CSBxpBdTCRUCwg/xsAwHBThrIFYoimTqvErtNkLPHAMCmHy2cfmkRVeUKeoPE5RLkZenxM7obx3WWjDFqfG1jEcTtghX/iuPHzwJZ/t5uwmPt6HRQclB9yGo0UFKgZeR4a+O5veZAFxUlGirLtM3mT0yz4m9xERTmojhfx09fWIgeau/QL73BzdAUW2OhJGePgePOK+lwnNsFCSOtxAx1kJdlIGunAXOga5DqMgDi8QXVJXJUiDz79TkddwRWjHvTm4LHd8AlUsqdQog7AP/6b5U0KXiESClvEELMA5byV8FjhZRyYnvzH2qpS2R9AvJd1FxfArCkvQdfgzUQKK4/LYnaaoXaKg2L545i8dxRXD4zjYpSLeXF2mbkitJCHedPHEleloHFs1JYPCuFr98PweWEWxaO4KQRY9n6sxkk3SJjnJqSzvHxY/joxVBy9xn4fGUoVWVaVj0Rwdp3wnjy1niqK3QU5xv45atAKsu0fPpGCItmqj69/GAUFSXaVvP//o2Fq+Ylk73TjwUZo9j0vcUrvwpz9dw4P4ncvYZGko034yrLtNy+cDh59STn7z4OGqS6DJB4fEF1UV97Fa8+Xtoy4HUhxBbgcNS3xQeAWUKIXaiyuQfq+64G9qIyA54HFnc0+aFe+ZU3JCvrvy6TUga3MdajyLlBYPzxK+HNBMwmfyfWWg1CoVHIGRCirloaCgYAoyZaeWjVLvbvUnWCQ4bbkEgKcw09Ijrd+ksAu7eZWHRHDhpdvUREAZ3ezU1nD2f8tCpOu6wIp0NgMLqpKtfisAmP/o8cX+OVwHjPNiPpU2pISreSn6Xnt28sjJ5c3eG4LevNfPNREM+t28mm78ysXOGdOHogioI9zdWSNOTLI0n/DiLniFGh8ozXjvOq79PjX++/ImchxH9R4QVJTZprhBA7hRBuIcQc/npfb2VtiZwbBca7/Hj+niFcfmwaV5+cSkCIk9jEumZCzu8/CWxWMADY/quZnRtN3LdoGEuPUyksOj09Jjrdn+nHsvv2c+sFI9jzhxFzgIvbL0igvFiLrVagKBJbjcK/zkpEq5U8sHhIm/7v2mL0yq/9uwysfCyCvH0GrjphBBNnVnhFdcne6cecs0tx1IlOXW8gioI9zdWSNOTLI0n/LiLnThQ8et16bOUnhDgasADvSin96tteQC1wZABbAauU8oaO5moQOTcVGNdWKzx3ZwxVZVoW35NLaKQDKcFhF5QW6hqF0CERjsac2auPRPL5WxEkjqrlmof3Ywl2EBDkoiBHj8HP3WOi02knV1BWrCE63sGX7wRysCKKfbssXHb1dgKC6tj5RzCv3RfYSsDc0v8v32kt5O6IzlJWpKPOKhCCDoXcodEOYhPUIz2dTrVIlDGtqtP3YSCIgts6prQpacjXR5L2dZFz+Kgweeor87zq+/yEV3p95deTIudFwLGAQQhxAPg/4AZgFTARdYubV2vkBpFzU4Fx6rhaPl8ZCsCebUaPYuXUcbW883QEd72yr/GISBTYu93EsnmpXLs8GyklUjanwfhadPr96kA+ejGMqx/KIW64nSOS9/DTd1Esu+AoTj9vDzPnHeCmp8px2AVXnTCC6gqtR/+nzisnKMzlFdXlyVvigM4JuRf88yBXnzwCk9nFGVcUsXubiaAwZ4/em7+rKLitY0qbkobAt0eS9nWRM9BnzufwxnrstVdKeQ7qQ+4PKWWclPIFKWWJlPJY4BfUQke30SdNySVOh/c3voFcsv03c3dd8Mo0GklSupVN35n5dm0MNquG+efvYfV7CVxyxgyevzuaOqvCZf/XLt2rU9aZezM8zUqdVWH3VpPPrj9oA8ukBIdb49WnL9ghLXg0aV8HXNee+tobqsu+P43c8mwWdTaFb/8XRMrYmlZ9Sgp0zc7HjUt0MGlmRb38xY9XH4lm8T37ezypbw5wExFrZ9P3ZrbtiMPpgPMu3UVZiYE1H8Wz6VMnVz2UQ1yinaI8XbcKHl2hughFBbS6XGCyuFEUOLDHgMnsQqujGVlmyuzKVm2DBQ/PP28tKTh9NR5fFDxCR4bL41862au+r01+oddfew9pwUMI8bAQYgdqzu9hIURQW+M7orrsz9Rz09PZfPNREOeOHUXahOpWSf39mYbG83EvnZbC9JPL2bzezC0LkvAPcLHuwxB2bTUdkqT+I9fEsXOTkcOOqGH/Pn/OunAP36yJ4aalkzl6Vi42q2Boch3bNpi6WfDoGtUlJNLJSw9GUXpQz+3nD2PD1xZKDurQaGUzsswnr4QhBJ2mzQyEgoenn7eW976vxuOLggcweHQltFnwmA18hbqtLRsokFLe2NFcnqguNqvgyOOqGoXCYdEOBFBZpmnsExlvJyLWgaKoR0Tu2Ghk12YTP34WwLNfZXJG2miGJtu6hGXvSgJab3Tjb3bjcgpcbrjkqFQW3Z3HuKOr0GolilZSVaZFyq4XPLpDdaks12G3Kax9L5TTLz/Izo3+RMRaGXPEX2SZlSsimTCjstNo+4FQ8PD081ZerP1bxOOLgkfIyHA558VTver75pTne33ld6gLHqVAFhCOKlqsRSUytGueqC7TTy3j0zdCWtFMWiZ/t/5kblUMsNs05Oz2o7ZKfVB0JUHc3QR0cYGe6gotD18Z3yZ5pSsFj+5QXTJmVJM2sZr/3BfHfYsSeWBlJpmb/ZqRZYAuoe0HQsHD08/b3yUenxU8BjH2asGjjZzf+wBCiP8Bb7U1vgXVpafcHLQm5hH73wo5MWiD5tmkFDj/Rg+/3ip43IKa9ztNeuHAsRkR8sn7opolbFui4KsqNEQOsbdK/kYPtVNaoGPGGWW4XYLXHolsddauJdhJZdlfmHlfJ6C7UiDo7I6LhvOEHXZBcb6Oi6eO5Pzr8znuvFI0WolWJ6koUe9VW9cLCHUREeOgvETLFbPSuPCGXIz+DibOqGrE2K96IpKMaZWdRtsPhILH38WHnip4BKdGyGkvzPeq7wdTn+r1195DXfC4WwiRg0pjtQDR3sy1aYuzVcK2JQo+eUxtq+Tv799YiB1mR2gkm743o9VKfloTCPx11u6l01LwtzTHzPsyAf37N5YuFQg6u+Oi4TxhvUFiMruJHFLH+/8Jo7xIy7/OSsTP5CZrh6Hd6/2wOpgHliUAEB5j58i5pUw6trIZxn5oipUl9+V2Gm0/EAoefxcfBnd4qHaoCx6nAfcAxwDnAKOklFd0NFdb5/Y2RcErGtkKR795vYXifB1XPphDbbWCJcDND58Fc9iUqsazdte+G8rRJ5YSFOKkvETbI6r/vduNXTr7tvWOi7B2Cx4N5wl/vzqYkeOrUTSwZlUYm38M4JanduFnclNZ1naMm380c+RxFQwbZePAbgOfvhHCOVcWIqEZxj4+yd6qrSO0/UAoePxdfOipgkdQaoQ86vmzvOr78dFP9PrK71AXPG4CDMAXQATqVrcOra1ze5ui4D3v8LAy4rBa/nnqCGKH2bj64QM8cOXwZnNPPb6UGaeW8PhNcXz1XnCPqP7feiKyS2ffttpxsdXUbsGjAaVvrdVwyXQV5HDB9Qe444VMFAX+ceRIKkq17e4MuffyBO56ZR+LZqkFlp/XBrS6nqe2jtD2A6Xg8Xfwoecw9n1nVeeNHfKcnxDiXuB81MOLpkspi9oY2wpj7wm53oCo3/S9Pwkjbc36+BklEklsoh1/iwtzoIuTkjMYf0wFi/5vP4pGYq1VCAm3U12pQaPgc6x4RYmWn78IaJYjG390FSMzasmvR0c1/DA2navkoJbwGAcOh4LR5KK2WsPLD0W1i7FvjtKvQ9FIPnsznLxsAxdcl0NgsIvqCk2b8VSUaNm50dQo0v3zdxMBIc5O34eBgH3vb/H4AmMfmBopj3jubK/6fnbMiv678qvP+Z0ENNs/JqW8RQhRCiwHrkfd79vK6rHWzwHMmhgh0yfXcPn0FBVQ8MufbP3Zn3eeCeeht/fyysNR7Nho5MSLSpr1KczTMiLdRmmhFkVxo9NL7nhhJ/Ej6rj5vGTGHFHFkruyAMHVJ45oHLd5vYkpc6qazeVN2xsb/8A/wM3FU1Mb+6z7KLAxR7bsgQNMP7mcHz8PQO/nxmRW/+NpQM3Pml/eONc727dhMEpKCrQoGknc8DpOuaSQEYfZ2vRBUeD5b3eQPqmKpcenUVygZ8VH23n7mSgiYh0U5eq4fEZKm/Gs+yiwUaR7+YwUnlu3g/BoJ5dO69x98Pbe9OS4vjpXX/DBU9sLv2yC7mLsJTjdf59qb09j7C9o2SiEGALMRkVMn+jtZJ4Q7y0R9S37FGQb2Pi9mQUZaTx1axwOu6AoV0vePh2x8dWcdkk+6z4MRtHIHsOKh4Q7ycvSU5SnBwTrPgzCYVcIi3I2Qx+1RM076hSKcnUsyEjj/kUJ5O7Rk73T2AmUvkJ4ZA3rPggkLaMCRSPJyzK0G09IuJPCA3qcDgWnQ2Hjd5YevTd/Z+x7f4vHVzZY8GiYXIipwJdNCh4jgPuBu1F3evwgpTypo3maYuwbMNwtEfX+AS6qyjQY/Zvj6IsL9PiZ3Iw/pgpzoAuHXZC52UhohBOdQaIzuPG3uKizKVSWansEKx6baMdgdBMY4qS2WsPeP/z48p3gZuijlqj58mIt/gEuLpuWymFTqrnizlxKC7WdQulPnFFFythaTBYXjjqBtUaDlG3HkzaxBv8AF1qtukPhwB49scPsg9j3ARCPLzD2lpQomfH0eV71XXfsv3v9tbcnpS4rgfeoL3gIIS5GPWdzGvAqKtLqZm/na4nhbomof/6uGHZvM7VCdQM8sHgo89NGs/yqIWz50Z8De/y46MiRLMgYxboPgsjLMnB6anqPYcU3/2jm3LFpzBs6hmduj2H/Lj+2/Gghd6+Bi44Y6RE1n7nZ2Oj/lvVm3n02vF0fPKH07740gRfvj2L7BhPr1wRy4ZT249n7hx9fvx/MvKFjWJAxiuxMv15HrvdV7Ht/i8cXGHtQhc7efPqCHbKChxDCBHwNzJZSVgghsoAMKWVxG2M7pLp0Bbl+8ICOaSdXcDBHj0YrsVkFIZHOQyo6bYmZX/1aKDs3mXxC/+irItrBePqODz0lcrakRMmxTy30qu93M5f365VfS5HzcGA0UCKEsANDge1CiChP4zuiunQVuf7r1xaqyjXc8Y8Elh0/gqHJtkMuOm2JmT/xwmKf0T/6qoh2MJ6+40OPiZylb3N+QogsIcRWIcQmIcSv9W0hQogvhBC76v8Orm8XQogVQojdQogtQohxHc5/iEXOdwDVUsrlHa38mponkXNnBcBNhc8N4miNxo3J7KayTIMQ7RNPfCk6bYmZryjV8MVbIcycX9ZvRbSD8fQdH3pK5GxOjpbpT7SqcXq0n+Y86M3RlVm0eEYIIR4CSpscXRkspbxRCHE86mlvDUdXPialnNTe/Ida5Nwl8yRy7qwAuKEtdZyVt5+KoM4qiElwsfE7MxWl2k6LQrsjOm2JmV/+/m6qyjX9WkQ7GE/f8aEnqS6HIJ93MmrdANQawjpUMtTJwCv1rICfhBBBQojo+qNyPdohxdjXf2tp/TmcX6EeXt5p6y5y3e0WLJ6dynkZo4gZZscc6OzSPN21Bsz8M7fHUGf7++ijBm3QPFkP7O2VwBohxG/1NQD46+xvgAIgsv7fsUBOk7EH6tvatEN9bm8kUIwa1N1AtJTyH22MbbPg0YBcr6sTBAarJfvcfXqsNUqrpG7c8Dqi4+2Ex6j6zdVvhPLETUO45NZcJs+qJDDUicFPlcT0FNWlFT0lxEVErB2DUfLKw1G891x4K0R9Z5LZoydXE5ugwlxdLtD7SZ67I4bNP5q586V9+Pm7URSV6iJE2zG2PHd26f05TJ5dOVggGADx+KLg4T8iWo5acZFXfX89/v7s+mdBgz1Xv7Gh6TMgVkqZK4SIQN0Suwz4yNPZ30KIj4EHpJTf17evBW5s76iMQ0p1kVIeRD1JfTswn3ZEzu0VPEIinTx81RBsNRoeujKeX7+24KgTrTD2v39jIeVwK7+ts3DNyUm4nIL1n1vwD3Dy+7cWlh6fzP5MPToDPUZ18TTu+08C2LnR1Eyg3bJg05lk9pBEOzedPYwz0kYhFAkSdm42cs0jOTgdghvPTMRodrP9N1O7c7U8dzYto5qQ8MECwUCIpxcw9sUNv9/1n+daziWlzK3/uxCVAzoROCiEiK5/xkTz19nfucCQJsPj6tvatENd8DgdNRc4D/UheLSUskPutaeCR162yup76/FIzriikMI8HVNmVzZL/m7/3cwJC4upqVLQ6SV6g+TmBSn889/76leQsP03fw4/srLHqC6exhXk6Jm3sLRRoA3w5dvBzDitvNvJ7OoKhYh4NzXVOqJja3j1hVQ2/x7Bsit/IW64jaI8fbtUl6bnzm781kzy4Z2/DwOhQNDf4vFFwcM0IkYmP3qxV303n3BPuwUPIYQ/oEgpq+r//QVwF2odoaRJwSNESnmDEGIesJS/Ch4rpJQT2/PhUBc8bgPcwAZUnP3l3kzUFsa+KE/V+21Zb2bR3QeoqdI0S/5OnFVNYa6e7Ew/ksfUYDI7ycs2sPT4tMa5px5fSsYxFT1GdWmr4HH+xJHc9co+Fs9KUf3w0fmuD7+7m5xsf2688iiWP/Ut+/cFEBpmxWZVcNgVLjpiZLt+NT13duq8chRtz55p/HctEPS3eHxX8OjuDI0WCbwvhAD1OfWGlPIzIcQGYFX9pols4Mz6/qtRH3y7UY/H6PD9+1AXPMpRd3dYgUDUpalHE0JcJoT4VQjxq6uL+62FkCSNruXjV8O5/YIkpISzFjcv/hw5twzphq/ea/Mgub+NaXVuktKt5OX6A/DoA2OZd+o+/nHFH2i1bpz2vqGsH7T+a77a4SGl3CulHFP/SZNS3lvfXiKlPFZKOUJKObPh7G+p2hIp5XApZXp7ub4GO9QFj22ouzyuBCagnuGR2BHK3hPG3pvkb+4+P+aeW0LBfgMarYqr37nRzNARVmprNPhbXASGOsjZZcBg6tlzezs6T9gX57v6m11otJL330nhuJP2YbXq1AKI3o3L6iAmwU5xvs5rv3771tLqLOTBAkH/jMcXBQ9jUoxM+velXvXddvJdA2qHB0AocBywEVgFxABhHc3lCWPvTfI3c4uJnF0G7rokkbsuGYY5wEVZkQ6AN1bEYK1VuOjodBJSe/7c3mY7PHrofNdtv5gwBbjI3mcB4L7bJrDs4hksvXga8cl1vPef0E755eks5MECQf+Mx2cFj0GqS5sFjyuAGCnl7fUPx9OBIG9Wfp4KHt4g5Dd8HcCDq/ZgMrvQaKGqTEEisFYpaHRQZxOERztwOASVpYem4NET57uazE6ihzrYst6f378L4PgFxXz6eiizzixDb1Crt7VVmk75lbtPR1yifbBAMADi8UXBw5gUI4ctv6zjjsCfp97Z6yu/Q13w+C/w3/rX32TgYm9Ob2ur4OENQn7vH0bmp41u3Elx+YxUnv5iJ1WVmkZJiEZz6JP6PXm+a+o4K0jB1HkVWKsVNv3g73WMTf1adPcBgsNdgwWCARCPzzD2fyOYaW8dXXk08O8OSt3tYuy9QXpXlWsYkW5F7yeJGVbHT2sCsNVqmDy7grJCHTecOZzHP8mkrEiLf6C72yjwvoIoH5ZqIzDM2WGM5SVa/IyShFQbJrMLRaue/NaA119yXw5zzi4jd4+aMz2w14De4CZtYi0Hc/S43eBvcZGz2zDgsO/9LR5fYOz9kmLl0Ie8EnCQefr/9d+VnyeMvRDicOAZ1DygVQgxUUr5i6fxHWHsvULIfxjITeckcvMz2Xz5djBHnVDBmlUWPnoxjOsf209FiY7fv/dn1hnlXHxU38KKd2eurT/7s+rJyA5i9GfsUdW8ujyKey5L4JwrC0gdX0NAsKsRry+EpDhfy6JZKWi0kpfX/8mOTUY0WskN84dTWarlrKUHCY+ra4bgHwjY9/4Wjy8w9shDsrfXZ3aoMfYPoW5rc6GCTB/ydrKuIL0RgsX35JGz24//vRyORifZn6nHUad+2+DnJPXwWqRb9DmseHfmqijRdhhjdaWWoDAnG38IQGsQHH1SJS89ENMMr5+WUYveAJHxdvxMbgJCnPywOhCNVhIea0erl0w7pTWC35fx9JV72t/j8ZlJLz99wA41xv5z4DfgSNQV4IlSynM7mscTxt4bpLeiwNR5FdTZBDqdpKJMw3/vjWb+okKi4h3kZ+vZu92IJdjZaex3X0aU222C5DG17cbolgpJabUERzjQ6yX7dvhx3WnDOfrE8mZ4/T3bjdz2fBZaneSPX/y57vQRvLlpG/4BKpPi168trH03uBmCfyBg3/tbPL7A2PsNj5VDHljkVd/dZ97W66+9ba78hBCP18MBPX46mrgNjP3VqBq/dNTT227y1tGuIL1LC3XMiT2ck4aP4YzRo9mfaWTnJhO3LhhOXpaBy6an8v3qwC5hv/syory2WtNhjEJIoofWcdO5Izhh+Fj++NXMWUsLW+H1f1kbyImJYzgjbTRut2BoipUlc5M5MXEMCyaMIireTvwI24DDvve3eHyBsZeoxCRvPn3B2lz5CSHapRJKKV/ucPLWIucVwDdSyneFEGcCl0kpZ7Yxtl2MvTcEki0/WTh+QTHWag1uNwSGOnn/uXC++ziIu17Zi9MpqLMq6AzuLgt5o4fWERTmoiRfR2GeziMFZcrsSrQ69X/qr94LImGklcmzqsjP1gMwJKmOzM1G9H5/+VFRqiE+uQ5HndI4LjzW7pVfJQd1HDalpvGc4JYC6rwsPbu2+jPnnBKEkJgD3NhqFUoParltYSLPrN2J3s+N2yXI2unHEzfHcdQJ5cw+sxSnU1CwX09QmBMBFBzQk7PLwMiM9sXQ/U0U3N/i8YXI2ZAYK+PuW+JV373n3NJ3V35SypebfoC3W3zdrrUhcr4IuE4IsRVYiLr9ra3rt4ux94ZAUpSrRVHghvlJXH1SMgXZBnJ2+wEQEOwiZ5cf1RVKl0ShDdSY5MNsFOXqEQoeffj9GwtCwCPXxoGQTDu5nFVPRFJRouXOfyRw9QlJKIpk5aMRLQSsNlxO0WzcQs01iQAAIABJREFU6tdCvfCrmgkzqlh+1RBy9+o9CqjTp9Swc5MJS4CTOquGs8ak43CAf6CLiDg7Gq3k2TtiKMjRs+qpCC6+JY+tP5nJ2W3g6dtiGJpiY9P3FpYeP4JR42vYv1M/4ETB/S0eX4mcpfTu0xesw5yfEGIK8AJgllLGCyHGAJdLKRd3MM6TyNkK/EtK+ZgQ4mFggZQyuiMnPYmcvSGQ7N/jxwkLS8jdpwcp+PZ/QQwZYePwI6sIDHVRWaahplJDbZWm06LQ/Cw9u7aZmDC9koBgF5YgJ2vfDSZ5TGtSyt7tRpbce4CwaCev/zuSlU/EMGFGBVfccQCjyYXTKbh/0VCW3JvbbGzJQR3pk6sbx236wdyqjyfha3xyHdYahcAQJzarwu6tfvhbZKtxh0+tJiDIRclBHRqtJCAcDHoXfkYnNTVaXE4NtmqJOUDlHX77cTDZmX7c8Og+7HWC0oM6dbvcABQF97d4fCFyNiTGyth7vFv57Tuv91d+3khdHgXmAB8BSCk31z/YOjJPImcXcL4Q4h+odBerN056Ejl7QyBJHWelpECHw66AhNIiLa8/msCtz+3jzccjMZldXHZ7XpdEoQ0o/XuvSGhE6f/+XQCKpqUPtbz1RCQ3zk/irlf2qcdUKrDhq0A2fBXItcuz2b3NSGiUw2OMTce1JL+0JXw1mt2tha/Z+lbj9v1pJHGklRvmD+eIOeVceGsxX36RwMf/G8F9D35NUFAdzjoti49LozBXRW9NPb6UilItL94f3SkKTn8TBfe3eHxDdek7x1J6Y15JXaSUOS2aOsTPt0F12QLcLaUcg0p3Ce+kv522a09NYuncFG5ZkMhJFxYzf/FByou1XUbgN1h3UfqgUlgmzy7n24+Du+VLV62mUsOqpyK4f+Ve5i8uprLCgNstmHfibla+PooDByw8e1c81zyU1TimP1FwBq0HrD9JXYQQ7wD/Bp5AhQRehXqi0tkdTt664JEKrEAFHHwEXCmlDG1jbIfn9nZEIMnKNDJ2ahXhMQ78LWppPz9Lj94oMZpcmMxuECDdUFqkw2ETXieIhSIJi3agaNQchtMh2PqTmcBQRys/92wzsvT+3MYCxKqnYsiYVsE1y/djDnTy+qPRbP3R1KooERDiImqIHT9/NycOO6wV6t4XSfCYBDtGfzd2q0LmnlA2/h7JwvP/wOUSaLVuivN1WIKc/POMUdz6zG7CIu243VBRqsXl8P5+9bcCQX+LxycFj2FxMvpO75Qy2Rf8f/bOOzyu4mrjv9mqblnFkmzJvcgNO7YxNiWfjXHoIZBQQg8dDDi0ADGBhFADJKEFQoDQDcShBTAdm24wLrjhXmTLsq3ey+6e74+5K69Wq90raVdayXue5z5azc6ce87s3dmZec+85+ZuX/aamfldBsxBJwMpBCYa/wcVpVQeMB8YrpRao5SaKyI/AmcApWjww+bNu+kvoQAPMwwkPy6L57qTR/CnC4dww6+GoRS89GAWZ08Zy+tPZvLui+l892kSInDPFXnt2iBOy3Lx/P3ZXH3cSIoL7az8Mpk//mZIQDvn3LWrBQAxaFQtc+4oYNOqeB67NY+ZJ5UFBCW+eKcP983dz8xtJjdxezfBX3kkk+tPGYozwUP//tV88XkeTU2K118bQUFBMs54DwDX/GUrVpuH8386HrtDqCqzHNAAQW/zJ1yAhw69NnN1v0SS1SUHmATcC0xHBzdfgD7yVgaMBhKAzSJyYzBdgQAPMwwkW9YlMPn/qgBITm2ib6aLE0ZqgPnsuTtJzWhiyKhaHE5PhwCPlV8lseiNvvxlwSa2r49j/kNZAQGJw46pRKAFAOFxWxg+ro7zpo/mxPOKyRzQGDC/b1q/JlIz3BTvtvPRgtSQuYnbuwk+cEQ9FqtgtcIHCzJ58OahvPL9UhKT3c2z2r2FTlLTm2isV1RW2MnqXw8K9u1qmxK/twMEvc2fsAAeQ3Il549Xmaq7/fybun3mFxLwUEoNBR4EpqFX618D14jIlhBN/4rOr5kBrAP2AWcCF6FTzP3H0LsInXezTWmL1SUUA0n+5Hou/9loAH7/2BbGTa1uoffgGRWgoKke5p4wguoKWzvZU2oBqK+1cut5Q9sEJJZ8nNKqrLjIwc1nDAOgeLdGdQPpX/BYP25/bmszqBMqN3F7N8GvPGYkWbmNPPDGRhob9ELg+tPGcsezP2K1CqkZTdx26RjEo7jjqTUk93Hhdiku/r989u5yHLAAQW/zJ1w09tGyn2dGzOz5fQM8il7Cgl62XhUqG7qfjsHAZ8A4YIc39ZzSBP1lvqnofNp0mtWlod5KVm4jfTObGDC0gcpSK1edNI4bHthCelYTbhesX5nIuIMr6ZPmpqLE1i5mjKpyK6Mn15I3rIHd23WukPgkd8h2Y6bUkJLmbmZP8T6M/vUccR4Gj2po3iss3OrkmDNLQvbDoJH1uJp0cPSrj2YG1O3bzuHwkNjHzaK3MhgxvobsvAaaGhVWq4cml5W0zEYsFqipsvLQH0Zw/Gk7yR7YQG21lcQUN3EJHrauiyM923XAsKD0Nn/CweriHJwrObddbaru9gtu7PaZn5k9vwQReV5EXMb1AhAXqpFSKk8p9alSah165vehiFQCdmMP0ANMpo3fCt89v+K9VsZPq+GWs4Zw8YxRjD+khupKS4uy3GH1pGc3tajjaoAnbs/l6bsH8OP3CST18ZCT18C/7hzIRbMO4renjGXiYZVk5Lgo22drU3dbZY31inuuGEThNieXzx7FoJF1TDg0dLuCzU42r97fhRk5TRTttPvVq2b8tJoWe4V2h8dEPzTQ0KCYd9aQILpbtrvp9GHEJXhobISbzhqNze5h+8Y43C4Lz/99IK5GCw11irh4D1abB5cLkvu6uPyoUbzzXDoFGx2MP6T9n4+Zfu5ou2jVFQ02BCo79BAbdJbVhZ4V5BzsbG+aUioNWKiUukkpNVgpNUgp9Tt0pqRQ4kIvZ3egU84doZQag86zeSl6JpjO/rybQaUjDBf19VY2rU5gxkllfPV+X5oaFDaHsGlNImmZjdTVWNmzy4nFqrnq2tLdVhno43EANrtgtbW2IVC77evjW7CnzDipNTOKzQ41FVb2FToAxaI3Uxl3SE1Iu9xN4IwTnPHu5pMhIVlXlM71q1DU1+qYyPpaK+UlduITPBRsiae02IHbraiqsDNkdD17dzqabS/Z49ARDAcQC0pv8yds4lHmLhOilLIqpZYbCclRSg1RSi1RSm1SSr2ilHIY5U7j/03G+4NN6Q9ytncrelYWyFIRkaEhDFfAs0CpiPxWKfUmOlzmZ0AJcAywCqgTkd8F09VRVpfyYhtP3pHDf9euoabKQp90NyVFNp67P5uDptcwamIN/Qc3UlVhpa7aikj7mDGGjtUbxmn93CCweU0c5SW2kO3Ki21sWBnfgj2lYLOzBTNKubF8ra/TJzVqq62UFNko2WMPqd/h9DBoVH3zyRB/3b7txkyuxRnvQVmgoc7CW8+kc/qcfYiA2w2uRsXTd+XgcsNVdxaybb0Tm0MPsPFJbiwKaqosFG5zkJHd1K7PpyezoPQ2f8LB6uIcnCs58+aaqrv9kt+FXPYqpa4FpgApInKCUupV4DUReVkp9TiwUkQeU0pdARwkIpcppc4AThaR00PZ0CbgISJDTHnRthyGPr+7Sim1Fn3G9zngHnTyoqlAPDqhUUjZV+hg3pkaIPDN2+stu/zPO0lMcbeisW+os/KrsWO5/7XNPHRjP75cqLcXF7+Zyv2vbea1JzJISvW02gwOdb9ZvyylvNjOo7fo7JuJKS7++sYm6mo0y0awdvmTapvZUy49cn/e3lD+eIGSUHblT6rlH7fktjgZ0pY/J7doV8fTdw3g6bsGkJji4rantvHovP7sKXDw9/9tZN9uO5fNykfZHS0+m9/9bSMDR9QH9bstfzryuZr9fKJRVzTYEKhs8C+/4Bcn7l/9dEjCGMCslMoFjgfuBK41JlNHokFT0BOrPwKPAScZrwEWAI8opVSoFBmmQl2UUuOAMfjs9YnIcyadSAIWA3eKyGs+5YuA69vKrxkqyHn3Dge5wxqoLrcxOL+e1AwNYFRV2FsECfcb0IgzXnjuvmxee0IfKJk6q5J5/9xGQ72Fz/6X2uH0jD98k8zm1fFcdvsurBahstxKelYT9XXdk7qyPawuZvwp2uFg7y47M35RTlOjBVej4oyJY5lyVB2X37YDi1V47+VMdmywcvOjO9hb6DhggoJ7mz9hCXIelCc5vzc587vshqAzP+Nwxd1ofoDrgfOBb0RkuPF+HrBQRMYZOYGOEZGdxnubgUNEpDiYDSEBD6XUbcDDxjUTzb78cxPt8owBrggdIJ1nlN+nlPoRPZ29TykV8JxUqCDnZYuTGTWxjg0r47ls1kj2FVpJTvX4BQmnsH55Ak0Nqnngs1g83PzYdha/lcqZPxnTqfSMxYVWrrq7gFvOHsqcY0cyaGQdaVndl7qyPawuZvwZPaWaI39ZzuK3Urn6uBEkprg1n9+ft/PwvEFcctQ4Zvy8hPxJ1didckAFBfc2f8IW5Gz+eFuGUmqpz9Wc9k0pdQKwV0S+77Q9QcRMqMsqYAKwXEQmKKWygBdEZHaIdjnAv4BNwB/QQc6/AHKBT4CPgO1AUUeCnNcuS+KEc4qpqbJgsRCQ1aWowMHx55TSWK8o2Kyn9KuXJHLSBSVsWRuHiEZDFVBZFjytY6Cy6korB8+soqjAgcUCDXWK9OzuSV3ZEVaXUP5YLB4GjmykusJCXIIHqw0aG2D3dic7N8WRO6yBlL4uGhsVddVW7A7PARMU3Nv8CUuQ86A8ybnR5MxvTtszP6XU3egtMxd6tZkCvI4mWMkWEZfBNvVHETnaYIj/o4h8rZSyoSdcmaGWvWZYXepExKOUcimlUtDobF6oRsAw9Jp9FTrYOQs4GVgLbEOTGkwEaulAkPPU2dXs3aVj64aOqaP/kEa2rovjgWsGNtfJn1TLuVNHc/tzW7li9v69NUectAwUPaLjgaKL/5faIV2RSl3ZHlYXM/4IrW0oLmqtK1BZbw4K7m3+hCXIWYAwsLqIyM0YLO9KqRno7bGzlFL/AX4FvIzOD/Sm0eQt4/+vjfc/MZMS10yc31Jjafov9OxtmXGTUA58ISJKRA5Cz/gqgIdF5HWD5cUJfA4ERXrbEqWE4eNqefv5TOYcM5rGBsXQMaYYsmISk5hESJSYuzooN6LBj03oMLmnjPKngHSj/FrgJnO2tiPQx4ifSRGRH9rRpi3AYx563++UQKN0KMBj46pEZp5cSnWFjZIiO0s+TOaCebsp27cf8GgLWLj3P5sp3OrE44H0rCaamlTUbUB3ZkM9NcPF3p36bHBKXze7tjpaUOS3pctMaoAYQNA7/QkL4DEwT/rf8FtTdbddfX30nvBQSk3yv4A0NBPLpFCKgwAef1ZKFaBTVyYDAZmcQwEePy5PRDyK4t12ANKzG0lIagl4BAIW6mosWG3CfXNzufq4EZTusbaino+GDeiOb6jXE5foYcE/M7li9kjKi61k9m80BeqYSQ0QAwh6pz/hY3XpORIsyPnTIO1ERI4MqrhtwCMfuAP4P+DXwBgRuSyYrkCAx8bVCUw8rIrEFDdp/VxsXhNHYrIHZ9z+TfdAwMLGHxI5+aJ91NXocX/550mtqOejYQO6M7pK9tgZM6UGixU+eDmNyTMqSevnQoSg7cykBogBBL3Tn7AAHgPzZMD15mZ+W+d2/8wvWJDzzE7qbgvw+A3gBD4E+qHprYJKIMDDn0L+g1f6MmVmVUhgIX+SZmCuqbKCPtXVYRr7aN3Mzp9U2xxwnJXbyMkX7+PbT1K4f+7AkO1CpQaIAQS905+wAR5RkpbSjJiise+IBAE8vCc90tGkpjMCtVdKXeKNAXL7nbfuLIW8L7X9wTOqSMvs9HnuqJS4BDd/eHIbH7zSF1dTz3koY9KDpTfR2Hf6Bm0AHsZ7NwNxInJbMB2zpvSTR+/KbqZc97ihsd5CVYUVh1NISHazd5edgcMbKdyqj14F2sA//co9WG0wbXYldqcHqxUaGyE13RXyVEZP2sz2P9my7vvWFPlmTp4ESg0QAwh6pz9hATzy8iT3mmtM1d1y3XXdvuyN2MyvLcDD5/3rgLuA00LpWvGDi6Gj63n23iyuP2UoVhv85eqBnDdtLHdfMYh1SxPZt8uOCDx2W06bG/gzTirnu0+TuO03g7l8dj6/PWk4uUPqTZ3K6Emb2f4nW8zQ35tNDRADCHqnP91wwqPbxcwJDwWcBQwVkduVUgPRUdbfhmjXFuDRBNQDTwIHA1+ISNDjcr6Ah90ppPR18/IjObzyj/4cNK2Smx7ezJN/zuGaB3ayb5cNlAq4gf/By2l8+mZfbntqKwA2m9An3UVlmRWlgoMBPWkz2/9kizPeg0WBxxPcRzOpAWIAQe/0JyyAR16e5M41OfO7oftnfmZOePwDnWP3SDQvXxXwX/TAFUzaAjwmAUcBe9BHV34fygBfwMNLuZ6erffpEpLdfPZ2Gg31FpoaYe6JI6kstQXcwAdAKeYcPYpH3ltP5oAmln+eREWprdcBHr4nWw4/vjwgGNSR1AAxgKB3+hMOwKOTAcxdLmZmfstEZJJSarmI/MQoW2nk3jV3k5Y09jOBI0VkrlJqGzoNZiv2hbZo7L2U66V7HKSkuYhPcPPrgyfyk+ml3PDQdkqL9J7fxlXxDMmvJzXDRWIfN1VlNsqLbVw6K7/5HokpLp5c/CNp/VzNe4XtoQJPTW8id3gj1RUaOa4os5KVGz6K8vbS0XvLKkpsTD+6kj5pLgo2O4lP9LBrq6Pb/QlEuT52ai17Chx4PJCY7KZgk7NH0L7HaOxbS1xunuReda2puptvurZHzPyalFJWjJW6UioTPRMMKgblzHNANjAYeAl9UPlRwKKU+j+gP3oW+LJ/exF5AngCYPbUfjJ+Wg2XzhyFxQL/+uxHtq6P44t307jpoS08tnANNocLZxw44j3MPX4ED7+7kVVLEvnoP3259altzDtrCDc8WNDiHglJHtwuQQTuvHwQ29bF8fy361j5dQLTj67i0pmjKN5tb7Ns3bJ49hTYuebnI4lPdPPC0jXEJQoXHp4ftJ2ZspeWr6WhQXHrOdpuLx397FPLQ+pa9KYmyikqcHDF7FGMObiau1/e0s3+rCExxeOnKxGrTfjdqcOoLLVx+pV7yMxtaOFj4HbhtKFrdUWDDYHKnvp2BYSBxr4nzfzMAB4PoRkV+iml7gS+QAMVoaQVjT0wG+iLXu6mGvf/i1IqO5Qyf8r1kkIb65c6KdgUx7mHjOaOS4YgAredN4TyYj2mlxTZqam0IqKoKtNlqf0gKU2hbHYu/3MhDXVWAMSz/1MzS/u9d5eTvplusrwJf2w6AXo4aMU7REfv21nNrxWo7vcnEOV6ebEdq03IHNCIzSHM+EVrH6OV9j1GY9+G9CbAA0AplQ/MQvf3xyKyzkSbtmjsDwOqReT+YMteX/HS2PtTrr/4t34c9avy5oDcGx7aQekefdzNSzOfmOwmNcNN6V4b4oE/nD+K6/66leQUF/FJbpZ9nsL02eWUFGmgpD1U4OXFNr79OIXL/rTLAGJc/Lgskcz+7aMQDwcdvT+9fs4gnXHN5YK13yXS1Ki61Z/AaQbsHHFCOYkpbgCWfprMx//t24rOPxpp32M09q0lbkCeDJxjbtm7cV73L3vNkJkORNNO/Q9NHVNjlIUSL439kQaN/bFoXi6AK5VSP6ADnfuYMXRfoYOTR43nuIETuH9uHh++2pfP3urbos6nr/flN4eO5jeHjuabD1JYtjiZucePZPv6OK45Uefl3fpjAtedkk9ZsZ3zjziILxam4XZroMTbznu/C48YHbLsu09SmHO0BlnefjadogJz7cyU7d3l4MZTh7Nri5P5D2WZbrdscQrnTB3DMbkTOWvyWLLzGnE3qW71Z8PKeOpqLa3qzDlmJCcOncDZB48he2AjA0fUm2oXThu6Ulc02BCobOEHjSa+hSakN838DDJTQc/64oAhwHoRGWvqBn5BzgYZarGh889AjohcEKBdUFaXH75OYvFbqTyxaD0rPk9i/kNZrQJ5fYN2S/facDiFbxelMv1nZSQmuSne7cDu8NA304XbrZOH79zsNE9j/3USW9fFh4USv61g5ey8RuISPZw45CBGT64xFaysg5ybmgGe3963nSkzqkzR6/v7M/1nFdjsepbxyWupZA5oDFtAbiva/J0OCjY6GT2lJmi7aAgKjgU5t5a4AXky6DJzM78Nt/aAmZ+IjBeRg4y/I9CJh0Ly+bUV5Cwie4Ar0KSmpwIntnHfoKwu29fHcfQZpTQ16J2PQIG8vkG7N50+jMQUNz98nczvzxpFweZ4zjt8AuceNgEU3H7RYC49cmS7AkV3bHCEjRI/cLByH+6buz823Gyw8jcfJnP7RYMBcMa720Gv7+9PDTa78MC1uc37ju++kB62gNzd221cdfdObjl7KFceN4Ixk2vYsd7RI4KCY0HOPV86dLxNKbVKRMaHqNNWkPNo4HJ0DOAVwE9F5ORgugKxuqTnNDFgcCMWq+ByKV57IoMpM6oDBu3a7BrRVQpqqy28/ND+vcLRk2u477+bjPy47QsUra9THHZsVVgo8QOVpWa6SOvXRGqGm+Lddj5akNrKx0DttqyN55CjKsge2MTu7Q5WL0k0xc7i709qhovt6+PIGdTQvO+44ssk5ty5KywBueVGSE7RdgdWm1C8295jgoJjQc6tJW5Angy61OTM77bun/mFDHUxcmd6xYIOUi40obutIOeL0aEy36Hp7C8NpagtVpffnjSimdVl06oEUjPcQYN2vQHAn73Vl6N+VQ5AenYTHy1I63CQ88KX9rftLCV+oGDlBY/14/bntjaDOoF8DNTulrOHtWhnlp3F1x+vrr9fn9ciDWY4A3IXv5XaoXbdHRQcC3IOID0syNlMqEuyz+UE3kHnyQwqbbG6AOXA80AdGuzIba/RnWV1iUlMYhIh6S2AhxHcfK+IXN/hG7QGPFYDnwJXo4/IvYI+Nyx+7doEPJRFyMhpwmLVcUquJsWqbxLpk+6KGIgQzg3oQGwzgYCEjubfDaePHbWhtwMEvc2fsAAe/fNk8MXmlr3rb+/+ZW8wGnubiLjRISvtliCsLunosJflwKvoUx4Z/u2DAR5pWS6evz+bq48bSXGhnZVfJvPH84dEFEQI5wZ0ILaZ1kBCx/Pvhs/HjtvQ2wGC3uZPOAAPBSiPuSsaJBiNvfdM72Powes/QI33fX9uvgDt2wI8fgr0F5FblVJPA78EUoOlmgsEeKz8KolFb/TlLws2sX19HPMfyjK1Ed9RECGcG9CB2Gb8gYTO5N8Nl4+dsaG3AwS9zZ9wAB7xOXky5EJzM791dwaf+Sml4tB8AE40NrFARG5TSg1BH4dNR48p54hIo1LKiT5OOxkoAU4XkW3BbDBztjfOUHYk++P9BAg6+NE24HEf8LSx/B0JXBgqx2YgwCN/Ui0A9bVWbj1vqOmN+I6CCOHcgA7ENuNvf2fy74bLx87mAO7NAEFv8ycsgAeEcz+vAU2AUq2UsgNfKKUWolNT/k1EXlZKPQ5cCDxm/C0TkeFKqTOAe4HTg90gGODRz0B6V6MHsNXAGuNvyOlxEBr7RhE5Gx3m8oOIPB+ofTAa+5jEJCZRKmECPERLtfGv3bgEPQlbYJQ/ix5bQIOwzxqvFwCzjCO2bUqwZe9u9IgaSIGIyO2hXQiat/cxYJOIPBBKh5fG3v80wrcfpfDAG5so3WvH4fTQ0KBQSgXd6F3zbSKHn1BB3rAGdm93sG5ZAkPH1nUp4OEPSACtQInObGZ3N0hxIAAEvc2fcAAe8Tl5MvR8c8vetfeEBjwMwPV7YDiaDeo+4BsjD5CXOWqhiIwzVpLHiMhO473NwCHBeAOCzfx2i8jtIvKnAFfIgS9I3t6JSqlvgIuAc5VSU0Pp8tLY+5/wOPu6PdTVWLjop/lcMnMk/Qc1htzo3bI2jnuuGEThNieXzx7F4Py6Lqdq9wckoDUo0fHN7O4HKQ4EgKC3+dMNNPYZ3pWdcV3SSpWIW0QmosPhpqLT3oZNgs38mslLO6S4bcDj7+iNzNnodfnvRGRGMF2BAI+0rCYGjmxAPFBWbOej/5jb1E/NdJGa4aJPmouyYhtWq1BT2bU09v6AxPMPZLFjQ1wrwKOjG+rdDVIcCABBb/MnLIBHdp4MO9fczG/Nfe0LdVFK3YqODb4RnUbDpZSaDvxRRI5WSr1vvP5aKWVDT7oyg+EJwQCPWWYNa0PaAjwETWA6Hx3kHPK0SCDAY+rsahobrGzfEMfQMXXk/6SOrevieOCaULlp63js1gE88t56+g9u7BYae39AAgIDHh3dUO9ukOJAAAh6mz/RBngYpMlNIlKulIpn/2TpU+BXaMT3POBNo8lbxv9fG+9/EgpIjXjqSmhFYz8AeB+9l2gBDhWR7QHaBKSx99JwoxSHH1dO8W47bpcFV5Ngswluz35K7zFTakhJc7N7m0ZVvYPfko9SuOz2XdhsQlyCpnjvk+7uNBW4WVrxihIb65cnNKeIXDg/jcKtTo45s6TLaNL9Kes3rYrnoEOrW/TX5X/eyfSjK2O07weAP+GgsY/PzpPhZ5ub+a1+IGSoy0FoAMOKHideNRKoDUUPfGnoWOGzRaTBCI15HvgJOh/4GSKyJZgNZkJdOiSBaOxFpNKAp+uARjSc/Qw6r0cLaYvG3kvD/fVHKYjA704bSfFuO898uYr0LDe/OWw/pfe2DU7iCj0kJBmbDDlNlOyxMefOndz862EU77Yzf/lq8n9Sy4VHdB2t+KK3+jSzzVx65Cgefncjr/4jE18fI02T7k9Z/+w369i8Nq5Ff1ntHtKzmrq0b3oK7Xtv8ydcNPbhmvmJyA/ogczuknvTAAAgAElEQVS/fAt6/8+/vB7NEmVaIpa3lwA09kqpMeg8vZcYrDD/AA41o8yfcjsx0U1dtRWb1YWrQSjdqxmcfetsXx9PRrarxUmKylIre3Y6KNruwGIRmhoVSnUtrXhapou9Ox24miy4miwsejOVcYfUdClNuj9lvc0ufP6/1Bb9NenwajxuFaN9PwD8CYuYBTt6wtneTilum8b+PWCWiCwyghH/LSLxwXR5aez9abgbGyxk9teUVfGJbmx2oaHO0lynvNjGhpXxLU5SNDYoTr9qLyVFdiwWqKtRpGe5upRWfOzUGhJT3NhsOj/u0kVJZGQ3dSlNuj9lfWKym3MOHsNhx5U399f65Qkk93V1iIK/t9O+9zZ/wkFjn5CVJ8PPNLfsXfX3KD7bGwZpi8Z+LfCkUmolegPT1Em/QDTclWU2rjp2JJcfNYolH6VQW21tVeeHr5LZtcXJbw4dzfyHstiz08FX7/Xh8qNGcemRo1i/IqHLacW3rInj09f7cvygCZw9ZQyrvk7qsK7OtPOlrP/rtXnUVltb9NcX7/bpMAV/d/jTE3RFgw2RpLH35u4NdUWDRBzwCMDqko/OCJeORmiuFpH0AO2C0tibCfjcvDqe8dNrGD6+jt3bHLz7QjrrVySELZjYn53lyrsLmPazyrCwunQmiHbgyAaaGixBqef9KesXPJ7J6//K5MG3N4YMAM8Z1Mgfzh7Kw+9twOEUVn6VeMAFBfc2f8IR5JyQlScjzjA38/vhoV4881NKxSmlvkOHsgwAvEnOG9AhLn3QG5cBEZlQNPZmAj53bHQy/8F+FG51MveEEZx4fnGrgN/OBIr6s7OMnVJNWmZ4WF06HkRbj9ulQlLP+1PWzzypnJyBDSEDwJctTmbAkEbOub6InZudpGc3maK2721Bwb3Nn24Icu52ifSe34vAXuAGdL7fucDN6Bi/V4F1wBIROTeYrkBBzmYCPlsENO+zU1Fq5cNX0jjq1LKwBIr6s7Ms/yzJFF28WVaXjtpVssfO+GnVQannA1HWN9YrHHESNAB85VdJFBc5uPreAqorrNRVW3ngmryQjDq9LSi4t/kTjiDnhH55MvJ0czO/lY90/8wvYqEu6D2/X6ODnJcCI4Dp6ODp0WgE+Bt08HNQCRTkbCbgM39SLY/O00TRWbmN3P/6JqrKrWELFPVnZzFLF2+W1aUzdt146vCQ1POtKetDB4DnT6plxIQ6rjt5fwqB9OymAy4ouLf5E21Bzl0hEVv2isgX6MHVgz7t8TA6CLFIREaKyEhgHnpJHFGJS3Dzhye38fit/WmojyTG0/PF41Fc8bN8zpoyhv5DGknq42rxflZuI+XFtlgKgZgElF5BZhrWmyiVCryOPuP7TCBWhgBtggIeZsCGzavjyZ9cy8gJtdTVWHnl4X6sX5HAeTfuprrcxuD8elIzmnC7oKrC3u5NY9+8wPsK7Xz/WbKpvL1mWV3MABeB9PsDPd8vTmbctOrQ/vjl0a0st5Ke1dQi36/dKQzKrycp2Y1S4BFY8UVSyBQCvQ0g6G3+hAXw6Jcn+b80t+xd/nj3L3sjDXh8a4S0fGkUTweylVKblFICjAV2BWofCvAwAzbs2OigvsbCey+lc8Fh+Zx4fjF1NRZGTaxjw8p4Lps1kn2FVpJTPR3aNPbNC3zxjFGm8/aaY3UxB1wEtMsP6Jl6VIUpu4oLrVx1dwG3nD2UOceODJjv9/UnMqittHLhT0cz7+yhNNZbeOrOnAMOIOht/oQF8IgFORuK9cFkh4jsUkolowe5PwHHoENcrgMWAd+KyD+C6WqLxj4U2FBU4OD4c0qbN/Wz8xp5/alMjj+7hJoqCxYLLP88mZETatu9Qbx72/68wBaLDlbetdVO7tDGsLC6mAUuQjLX7LPTULf/FEuwdtWVVg6eWUVRgQOLBRrqFOnZrTfUcwY1Mu/MYRw0vZqr7ylgxVdJBxxA0Nv8CQvgkZkn+aeYnPk90f0zv0gCHjnAswYhoRWoRiO+r6MPJg8A+gJPhVTUBo19KLAhf1ItR/fXETZewOP7xX04eGZVMxtM/yENpthg2irzzQt8+Z930jfTHRZWl/YAF4Ha+QI9Dy/cwLefpHD/3NA+Lv5fagt/AuUh9iZ4/+HrJOY/lHVAAgS9zZ9wAB6K6AlgNiMR3fPzZ2IVkRt93tsGTGmLadV3z+/j/+RNnj4lvgU7yxlX7SU1w9W8b3b48eUtWFE2ropnSH49KWluKkqtNNZbmP9gP0qLHfztjY0UbnPialRY7R4sqiUbTGdYNgaNrMfVpPfpXn00k5knl5EzqJGkPp7mMn/bgTbtD+VjILsccR5yhzViswl2h4dPX09l8Oj6GAtKFOiKBhsixeqSmJkno0+6xlTd75+6rttnfhHd80NzaymgADhNKTVOKfWiUmo9OmXlg0Zyklbiu+f31v+a2Lw6rvm9jJwmVi1JbLFvVrrXxvhpNdxy1hAunjGK8YfUsPXHOP5w7mD6DWjik9f68uXCVIp2ONm7y84FR4zl8qNHY7MJ2QObWrSrrrS00mWmLHdYAw0NinlnDWm2s6hALyd9y/xtD2Z/KB9b21XN+Gk13HbeYAq32VEKNq+J76A/9aRnh6tvOqYrGmw4EPw59BAbdJbVpYft+UUy7sObfWkCcBA6Ackl6MDnfPTJjzg0nX1QefbFhlbsLG8+lUlNpbVFPX+WipIiO6fP2UtTg+K1JzIBKNvnYN8uB7lD67HZhIQkDyLhYcZwN4EzTnDGu5tBineey6BPmrtFWSDb27I/lI/+NtjsUFNh5cy5e9mxIZ43n85oxRhj1p9oYCCJBhsOBH/CJbGzveDPxJqGpra6VUT+ary/DXgSiBeRecF0pag0ueX0iS1ORAzKr2fiYVXNoME3HyaTM6ixBXOFxSIcfnwljfWKgs1OAP59T39K91n56+ubcDiFsn02CjY6ychpam7XUWaMTavicTg9DBpV3wxSzH8oi4OPrGTOnTuby/xtf/6BLGoqrRx7VkkLXUPH1pGY7A7qYyCmj7R+TWQPbGLL2jgSU9wk9XFTuNVBfKKnXf5EAwNJNNhwIPgTDlaXxIw8GXOiuWXv0md68bIXDXh8qpSqA/YBy0Xkr0qpq5VSO9FJSW5FzwpDij87yz1XDGLu8SN1IqMpY1jxRXIr5orSvXbOnTqawm1Orpg9iitmj+K7T1LYvCqRk4ZP4Fdjx1FVbqOqojUbTEfZTPbucnDjqcPZtcXJ/If04ZXvPklpUeZv+/vz0wPqWrY4OaSP/u02rIynod7C0f0ncPlRo3j23mzWfpfAptUJPZKBJBpsOBD8ibG6ROom+4OcrxKR1UbZv4AaEfltG22aAY84Eib/8dRpLWjfX30ki6N+VRo0wNg/CHnh/DRefTSbKTMqgwbydjRQNFDayFcfyWLKjEquvHtXizJf2xfOTwuYXrIjKSgrSq0k99XchnXVVlLSXFitwt5CeywoOAp0RYMNkQpyTszIk7HHm5v5ffdcL575tRHkfIxS6imlVBGa0TnXoLxqJb6Ah9PSMph45knlDBpVFzLA2D8IeeZJ5YydWhUykLejgaKB0kYOGlXHnLt2tSrzt6t1esmOpaAcOraehCQPVqvw+18PoaLYxs2/HtrtQbS9LSi4t/kTriDn2PE2ggY5NwBnArOAO4G9InJPMF0fvTJYDp/mbA4m/nF5PENG14cMMPYPQv5xeTwetzIVyNvRoNNAaSMHDm9EoLmssUGRlOJpYVd5sS1sKShL9to58uQyinfbeW9+Oiu+TGbOHTtjQcFRoCsabIhUkHNSep6MO9bczG/Ji90/8+uOIOcvge3oMJhcYBkQdPB75vl6GnZltwjSjIv3mAow9q3TnkDejgadBkobueTjlFbtiouCp5LsTArK/El1FO92UFVh5bBjK8jo39jtQbS9LSi4t/kTPlaX8EymfBKgZaGDY54QkQcN8PQVdFK0bcBpIlJmUOg9CBwH1ALni8iyYPeIJKvLD8AUNKvLIOB5EVkiIjZ0GsscNN3VLyJlw4Es1548nCuPGcW8s4dy8Iwq0jI7n5grJjEJJWEEPFzAdSIyBpgGzDESoN0EfCwiI4CPjf9Bp8kYYVyXAI+FtrX7AA8rmubqOxH5d4A2naaxD0QXf9SppR2ing9UVlZs5ZCjqinbZwOBqgorWXmN3UpjX2gse8dMqcVqERbOT6f/4HqmzKhqN6gTDZvznbGho59PtPoT7YBHUlqejD86IH7ZSr55+fp2LXt9EqA9AswQkd1KqRxgkYiMUkr903g936i/3luvLZ1dDnj4VPkbcD7wy0Dtw0FjH4gu/tVH+3WIej5Q2Q9fJdFQr/jDOUO47uRhjJxQa2oDOpI09uOnVzP1yCruuGQwF8/M58iTyxhxUE2HQJ1o2JzvjA0d/Xyi1Z+oBzxoF+CRoZRa6nNdEmQsGYzO4bsEyPIZ0IrYT4Y8AH2SzCs7CcEV2tWAx+3AG0AqmtL+dOAhEbk+mK6O0tgHoouf/1AWs08raTf1fFt09FvWxnPZn3ZhsYLFKribVEj2lEjS2FeWWknNcFFbpU+G9O3XxJ4CB1YbBxxA0NHPJ1r9iXrAIy1PJswyN/P7aoG5mV+ABGjlIpLq836ZiPRVSr0N3GOQKKOU+hi4UUSWtqW7OwCPZ9HJjHaiT+jcHlJRJ2js/eniYX/AdHuo59tiT3nlkSy++ySlXewpkaSx9wI4l88e1aJOKIAlWjfnO0vn35HPJ1r9iXrAQwjreTnj3P9/gRdF5DWjeI9SKsdn2bvXKN8F+B6Yz6UNrlCvdDXg8Q06cdEfRCQfaBCRykDtlVKXeKfE7k6et460eGnyP3ilL64mFbpBTLpUYp9P10m4AA8DvX0KWOc9EmvIW8B5xuvzgDd9ys9VWqYBFcH2+7StXQt43Abchd6IdCmlqkUkYJCzr8ya0k8evSu73Zu/AU94mDgZ0h4ae9/ct5/9L9WUrkjmEzbbzgwYdP2D2zjihEr2FDhAYPcOB/sK7T0GIOjo5xOt/kQ94NE3TybOnGuq7pev3xB02auUOhz4HB0R4g2L/j163+9VYCA6ZO40ESk1BstH0LhCLfCbYEte6HrAYyYwGahTSjUCiUqpHaF0rfjB1bG8vQFOeJg5GWKexr5l7luzuiKZT9hsOzNg0MiD6qirtjYDBpN+Wt2jAIKOfj7R6k+0Ax5eMtNwzPxE5AsRUSJykIhMNK53RaRERGaJyAgROUpESo36IiJzRGSYiIwPNfBB95zwGA+8LSIL2jPz6wjgEeiEh5mTIWY3jf1z32bkNKGAyjJr0HaRzCdstp0ZMOiDl9PYtHo/YOBqUrzzXHoLO6MZIOjo5xOt/kQ74JGcmis/+T9zM7/P3/rdAXnCY3y7FXUQ8Ah0wsPsyRCzm8a+uW/NnhaJZD5hs+3MgEFe8QIGgeyMdoCgI59PNPsT1YAHRA1RqRnpchp7pdQz6CxuDRgR2iLSEKBtc5DzGcenTf7tWYN6NK346Mm1xCd6qK+x8PF/+/LovFwychr416IN1NVaqCixUV5sJXOAq0v8qSixsX55QtD90IXz09iyOp4r7thF5oAmPv9fH756vw+/uWl3C1r+mSeXRayfo5X2vbf5Ew4a++TUXJl0hLmZ32dv9+KZn0Fj/xm0oLF/HrgZuBLN6vIbYCRwvH97EXkCeAJg9tR+Mn5aDZfOHKXJP79dx8qvE5h+dFVz2UvL15CY4uHCw/PbrBPOdu3V9ddr8igqcPDXNzfhalJYbcIDr29m1ZJEbj13KFfeU8Cxvy7lwiPyu8SfRW/1ad77vPTIUTz87kaWfJjiV7aBn59fwu4ddt59MZ1Zp5Tx7ScpVFdaufuKQdzwYIGm6t9pZ/ap5T368+kqXdFgQ6Cyp75dAeGgsff0nKlfd9DYH4OOxxmFPtcbMIeHv/R0WvEflydQVW5tLr/2gQISUzzNy46lnyRjsXadP2mZLvbudOBqsuBqsrDozVROPL+4RdmapYkkJrvZvCqBBY/1Y9GbqWQPbGxFy7/2u6Qe//nEaOzDJGLyigLpchp74BzgVGAz+ohbvYjc1LYmOPuEbLn67P49nlbc7hSSUtws/zKFw48tx+2GbT/GA/DtR0mcesU+inY4usSfsVNrSExxY7Np4GfpoiSmzqpqUVa2z8rICfXNgEFSHzdb18bx9nMZLWj5CzY7W1Hw98TPpyt0RYMNkaKxT+6TK5MPvdpU3cXv3djty94up7EHxqGBj1r0IDjfjLLeQCt+wynDqa22ULrHzjEDJ1NXbWXOsWOYc+wYNq1OwONRXebPljVxfPp6X44fNIGzp4xh1ddJrcrWLk1k+wYnlx+lUwA8e282e3c5WtHy95bPJ0Zj33lRHjF1RYN0OasL8A1wm4g8oJQ6BbhGRI4I0KbTrC5mgns7w+pi1obpP6vEZofqCiuJKW7eeLofm1Yn8MenNlOyx867L2ayba2deU9sY+8uZ1QH0f7wdRKbV8e3oOU3S68fjf7EgpzDF+ScnJIrU6ZdZaruog9v6r0zvyCsLhbgQqXUCnRE9rRA7cPB6mImuLczrC5mbFi2OBml4IFrc0EJiclulnzchzl37OCj/6Tx3vwMZvy8lIlHVGK3E/VBtDs2OFvR8puj149Of2JBzuEOchZTVzRIdwQ5ZwIbRORppdRiIE9EhgbT1dEgZzPBvZ1hdTHLLLNlbTzzHt9GXKIgHqgotVJRauP6U4Yz7/HtDB1bR2O9orLMhjPOE/VBtIcdU9mClt8MvX40+xMLcg5PkHNKSq5MOdjcluGnn9zc7TO/7ghyXg+8qJS6Do34/jSkok4EOYcK7u0Mq4tZZplXHsni0iPzuf25rc33nDKziqoyGzedPiwg80o0B9EGouUPRa8fzf7EgpzDF+QcLbM6M9IdNPblInI8cC/whoh8FSkbYhKTmHShmA1ziZLxsTtp7BcCT4rIf9toExTw6Chw4c+o8v3iZMZNq44Y00cMIIj5E202RArwSEkeIFMnzTFV9+PP5nX7src78vbOMspmA9cqpYYHah8K8OgocOHPqDL1qIqIMn3EAIKYP9FmQyRp7BExd0WBdAfgcSnwEjAEzc01VUTOD6YrEODRUeDCn1GloU6hFBGlNo8BBDF/osmGiAEeSQPkkIlXmKr70Ze3dPvMrzsAj0uAE4F56KQkhSEVBQA8Ogpc+DOqdAW1eQwgiPkTTTb0hLy9XSFdDngAF6ETDj+JPuoWMGF5pGnsY9TmMYlJBCQGePjdpOUJj9uBe0VkiVLqBmCUiFwUrH0gGnt/inqzdPSbV8eTP7mWkRNqqauxsuiN1IgCHr4nPMqLbXzyWmoruvhw5u2tKLUycGQDTQ2W5vv56+5Jm/PRYMOB4E9YAI+kATJt3KWm6n645LZuX/Z2NeBxLPpEx2NKqdXAIcChoXQForH3p6g3TyHvoL7GwnsvpXPBYfkRBzz8T3gEoosPZ97eoWPrcbtUi/v56+5Jm/PRYMOB4E9YAA9Br/PMXFEg3ZG3917gWBH5QCn1FjBARCYH0xUI8PCnqDdLR19U4OD4c0qbmUr6D26gpspKfU1kNqC9Jzx8WVAC0cWHK2/v7m0OSvbYGT+tuvl+/rp70uZ8NNhwIPgTDsCjT2J/mTbG3Mzvg6V/7PaZX1cDHp+jGV4eUEp5fwPKQypq44RHR+jo8yfVcnT/CcB+wGP558kRAzy8JzxuPHU4tz+3tZkFxZ8uPlx5e7339L2fv+6etDkfDTYcCP5EG+ChlHoaOAHYKyLjjLI04BU0ZrANnbmtzMjc9iBwHJot6nwRWRbyHl1JYw/cZBj9SxFZqpR6EE142iqvR6Ro7CtKbCz5MKWZmv3HZfE01FvaTftt1gbv/XyDnAPRxRdudXLMmSVhoSh3xHkYPKqh+X7+utui11+7NIHiIgdP3ZHD7x/fzsgJtdidHop3O1AQo33vxf6Eg8a+T2J/mZZ/sam6Hyy7PVTqyp+iJ0zP+Qx+fwFKReQepdRNQF8jNcZxaDzhOPRW2oMickgoGyLJ54eIuNGpKrcAFwJjgWuAT5RS9cBRtLED4BvkXLzXyvhpNdxy1hAunjGK8YfUUF1paVGWO6ye9OymoHXGH1KDywVX3r2Lkr02nv1LNunZTUyYHly32bJANrhctApy9k+fOfOkcuwOT4f8aV1WzfhpNS3u56/bt91Dv8vlmp8Px5ngobrSSmb/Rk67ci/Lv0ji0zf7YLPDii+SItI3kWwXrbqiwYZAZYceYoOw0NibvEKpEvkMKPUrPgl41nj9LJoJ3lv+nJG+8hsgVSmVE+oeER38DJmL3kgtQVNanQFcIiJxwFbAZUZJOOnbHU5PMzX78s+TsVglYrTiaZkuCrc52FfoAFRAuvhFb6Yy7pCasFCU2+xQU2FtcT9/3b7tfOn1y4vtDBjSyBHHV/Dp66lMOrwat0sx6afVEembnkz73tv8CZcoj8fUBWR4Q9mM6xIT6rNEZLfxugjIMl4PQOcJ8spOoyy4rREGPDKBh4H7gBeB89AjdhZ6H/ALABE5OJiucNLYZ/ZvZPTkumbAIz27CUQTjQZr1xla8QFDG3HGe+iT5qK22orFItgc0oJCPiO7KWw06YNG1lNfZ2m+X0mRjZI99jbbeen1P34tne8+7cOfnt5E0U4H65cnkpzaxKQjqikqcMRo33uxP+Ggse+TkCPTh19oqu77q+4MCXgopQajc3x7l73lIpLq836ZiPRVSr0N3CMiXxjlHwM3hkpcHmka+y+BgcBjQDGaxdkCrAJ+AN4B4s0oCxeld3WFlYUvpTVTsy96I5WKUltEacVXfpXEmT8Zy/GDJvD4rf1Z931CKwr5jvoTqF1xkb3F/Yp324O289LrA3z3aR9qq61ccMR4vnyvL/sKHdRUWWO0773cn7DQ2AuRPtu7x7ucNf7uNcp3oZOieSXXKAsqkZz5nQAcJyJXKKVmANcD5wPfiMhwo04esNA7svu1jwiN/e4dDnKHNVBdbmNwfj2pGU24XVBVYY+qoNPOBNGmZujMbKmZLlL6utm11YEjztNmO4fDQ3JfN688mt2CXn/ppymMmVJF3vBG9u2yx4KCe7E/4Qhy7hOfI9OHXmCq7vtr7+rIzO8+oMQH8EgTkd8ppY5Hp8P1Ah4PicjUUDZEcuZ3GPBzpdQ24EN0gPODQLZSapNSStAASMAROlI09ssWJzNqYh0bVsZz2ayR7Cu0kpzqibqg084EOccleljwz0yumD1SJ0Lv3xg0kHve2UOIS/CwY0NcC3r9MQdXM3hUA5+/kxILCu7l/oSL1SVcNPZKqfnA18AopdROpdSF6KOws5VSG9Fgqfdo7LtoUHUT8C/AFLtCxI+3KaWuRY/I40UkSyn1IfAWcB2wCPhWRP4RTEc4aezXLkvihHOKqamyYLHA8s+TGTmhtt1BodEcwFqyx86YKTVYrPDBy2lMnlFJWj9XQOaaAUMacMZ7UBaor1Ps3eloptcfMaEWh1Mo3q1TK8eCgnuvP2EJco7PkUMHn2+q7ns/3tOrg5xRSuUCxwNvA95YvkuBl9FoTF/gqVB6wkljP3V2NXt3Odi+IY6hY+roP6SBrevieOCayLG6dLWu/Em1XHjEaEAHcp988b42mWuuPGZki7LiIkczvf7lf97JT46IBQUfCP6EJchZBNxRcnbNhEQ61OXvwO+A5cB3ACKyxViP7wIuFJGGCNvQQpQSho+r5e3nM5lzzGgaGywMHVPXlSZ0mcSYa2LS5RIjMw0MeIjICT7vbwOmiEhxG+0jQmO/cVUiR/+6hJpKK1YrbF0Xx4TDqijb1zsAj8JtDlLS3PQb0IgzXnjuvmzWfZ8Qkja/M316IAIEvc2fsAAecdlyaO45puq+t/n+bl/2dingoZR6QSn1olJqPdAfeFApZQ/UOFI09utXJJCU7OaR3+dx8czRjJlSTUJS7wE8xk+v4Yt3Uli/PIGmBsVrT2Sy8Yd4U+26OqdxT+rT3u5P+FhdxNwVBdIdgMdxwEL06Y7vgY9E5LFgOsJJY1+43cGwsfXY7ILNLsQnuXE6hdrq6NqA7owuL3NNY72iYLMTZ7wHiwKPJzhVf1fnNO5Jfdrb/QkL4OHMkkP7n2Wq7nvb/tbtM7/uADyGo4+iZKMBj6HoIOg2JZw09jNPLmP7hriWm8bdvKkfCcDj3Kmjuf25rVwxe3+eYDPtujKncU/q097uT3gAD3oU4BFpVpcFwN1AMq33/OzoBEZzReTzAG2DsrpUlNhYvzyhmRll3bIEUtJcLeqMmVJDSpqb3dv0l9n7ofcElo1Is7qkpjeRO7yR6goriD7yVFNlbdWnoybWktTHQ3mxjVcfzWzVf9HMguLvY0WZlazcGKtLxFhdHFlyaNYZpuq+t/Oh3jvzMwCPvcAK4EcgxSh/Cp3bIxfN2rA8UHsReQJ4AmD21H4yfloNl84cRfFuO89/u45Fb/VpZka59MhRPLHoRzJzXFw8Y3+dbRucxBV6SEjSA3xGThNFO+3MPrW8WddLy9eQmOLhwsPzm9ut/DqB6UdXtbifmbLo0LUWR7xw27lDuOqencw8qZxX/5GJf/+tWxbPngI71/x8JPGJbp79Zh0fv5bq16fraahXzDtrCDc8WBCw/yLtT2f61N/HF5auIS5RuvXziY5npHXZU9+ugM6yuugvbqdVdJVEHPBAk5fmolkcXkBTWr2GJjb9H/pYSkgJxM7iy4wSiJ1l+/p4MrJdLeji136X1CNYNiLN6rJ3l5O+mW6y8hpoalTY7EJjnaVFn37wShoZ2S6c8e5mSnz//ou0P53pU38frTYQT4zVpa2yzovJMJcoGSAjvezNRbO4vA3cZAAeFwEXALOAvwLbROTeYHoCsbqMnVpDYoq7mRll52YHA4Y0tmC4KC+2sWFlfAu6+ILNTo49qyTqWTYizR7dM7QAABRaSURBVOpSXmzj249TuOxPu7A7hcRkNzs26cBvX7aZvhkuBo2qb6bE9++/aGZB8fcxpa+LH5clktm/fbZHiz9Rz+pi7yeHZpxqqu57Rf/o9mVvVwU5+46wjwPjgErgbCDJjCJ/Voota+JaMKNs3xAXkOHCSxf/m0NHN1PI9wSWja5gdfnukxTmHD2SylIbf702j/8+ntmKbWbvLgc3njqcXVucAfsv2llQfH18+9l0igrab3s0+RPVrC4Qm/mBqSBnK5rr7zsR+XeA9iFZXX52ehmFW514PJCe1URTk2oR8Omf3nLh/DRWfZ3UIuA3WoNOIx1E+8PXSWxdF8+8f26jod7CgsczW/VNoXFO+KDpNc3gSaA60eCPGR8/+1+qqfSm0epP1Ac52zPl0NRfmqr7XvE/e/XML2CQs8/7f0NTXAXsrVBBzssWJ2O1CffNzeXq40ZQusfaKiDXP73lzJPKqa9TPSLoNNJBtDs2OLj5se0sfiuVM38yJmDfjJ9ezcFHVrWgxG9dJzr8MeOj2fSm0epPTwhyFvGYuqJBujTIGR3bNwxIRdPbn47m3ro+mI62gpxPvriYumoFSgUMyPVPb/njcr0P1BOCTiMdRFtfpzjs2KpmRuvUDBebVsWT1s/VQtfAkQ3U1ej9w/o6C5tWxZGYLFHnjxkfM3KaUEBlmTXqP58eGeRsy5TpKb8IXRF4v+zJbp/5dXWQs0IDIBPQPPsKncs3qAQOcq6joc5CTbWO4RIImd6yJwWddkUQ7cKX0kLqik/ytNa13RGV/oTyMRoC2qPhGekJqSu7QiI6+LEf8EhG7+15lFKvAgtE5G9KqWoRqeyo8mtPHk5JkYM+6U08/M4Gtv0YFy67YxKTmLRXRPQZyh4iXU1jfwnwKjBDRFzG4BcQ7Q0FeGjmkibKi21cOiuf3963nSkzqqiv6x0b0L1tQz3mT/TYEDHAw5oh0xNPNFX3/apnun3Z29U09mvQeXzrlFKNQKJSakegxqEAj28+TOb2iwYD4Ix3M2hkHWlZvWcDurdtqMf8iR4bIkdjL4jbbeqKBukOVpdn0ElJFgSb+flKIMBjy9p4DjmqguyBTeze7mD1ksQOMZBE6wZ0b9tQj/kTPTZEDPCwpMs053Gm6n5Q/0K3z/y6g9Wl3dIWq8stZw/j9ue2doqBJFo3oHvbhnrMn+ixIbKAR/j2/JRSx6CTnlmBJ0XknhBN2iVdTmNvyJ1KqR+AfymlnIEaKqUu8WZzd4fhvHVMYhKTyIoA4hFTVygxDkE8it4uGwP8Wik1Jpz2Rmzw87K6iMj3fm/dDOQDBwNpwI2B2vvu+e0t0ukXvZKR09ScUcwrJUX2VnWKChwhy6w2wWqTdreLVl3RYEPMn+i2IVBZ/xwLmEj0HVREDOYIE1domQpsMnL+NKKTnp3UKfv8pKsAj5eBI5VSL4jIbtHSAPwb7WRQWbbCxYAhjWTlNTSzs3zzQZ8WddavSGhV553nMkKWjZ1SgzNO2t0uWnVFgw0xf6LbhkBlvzzJCTqlbKckjIDHADTpsVd2GmVhk4jt+YnIzehZHj5ne89WSuWIyG6llAJ+gQmEqdxdVnzNXU3ld71aM9huxz3/9arin88rjj9kUnxyah+L/bmV3zfd+0hZ4TV3uZruXVCTB1hfeatqz7/Wrioqvyupj2+7QGXLfmiou3dBbat20aorGmyI+dO7/Hn6pUr3QbNK1nTmO19F2fsfyYIMk9XjlFJLff5/wuDw7DoRkYhfwAw0wgvwCbAKPei9ACS1Q8/SjpR1tF206ooGG2L+RLcNndEVDRcwHXjf5/+bgZvDeY9In/AAQEQWAYuM10d2xT1jEpOY9Gj5DhihlBqC3os8AzgznDfoksEvJjGJSUzaI6JPgF0JvI8OdXlaRDq1LPeXnjb4BdoTMFPW0XbRqisabAinrmiwIZy6osGGzuiKChGRd4F3I6U/4ic8YhKTmMQkGiXSQc4xiUlMYhKd0t2oTjvQn2OA9cAm4CbgaXRqzNXG+3nAp8BaNIHCXCAO+BZYaZT9yUefFX3yxItCb0Oj0CuApWjC1QXotJvr0MSrK3yuSuC36Gx0a9Do9XzgOqAMcAG7fe73AtAINKCJHvqikW8XOjh+ilFvlVFWD7xutKsx/l8BfGDcZy+w22ibYfjiBup86u0FCg0f1gDl6PSEdYa/K4A3fMqWAicYfVZr3HM1MBqdba/asP9HdB6W1UYdAf5p9H+xUWcX8B6ww6hTZ7Sb4vM5FRltDwK2+tix2/jM1hqv9xiffY2PrhLDp+8MW+uMex1u9KG37gbDr3VGvVLgP8B2w04B/oJ+VoqNNvVGfyahsw/WGWXrDF3e52mJ0Sf7DF11hk9TjT7cY5TvM+5X7VOnEKjys304cCSwzCgrQG9NDTP8qQFeQT/bmwzb3zeemxeNPvLqsqO/I3XoZ3WB4Y8V/Vy4jHbPGH3vfa4ndvd3vauuHjHzC3TUBf0FOsanmgu4TkTGANOAOcBQ4EgRmQBMBI5RSk0z6s9FP8y+MlNEJoo+cP0g8J6I5KPJV9833puIZqapBb4CrkYPXOPQA9o1wGloqD5ZKTXc0G1BJ2/aCHyMHsD/DZxq6PLKM0bbTegvrjL83GTc+20gATgP/TB7WXFWAA8Bm416dwC3oPMlTxCRscC5wCHAZuC/6BSi/Q2bNwO3Ar8HEo2+zkQPrDcBWwydGUA6OvTgVqNvvkCf4X4Ejchlob+oJcDnIhIH9EOf6PkD+gfiaPTg6QIGoQfrm416tWiS2znoAbMSuBj4uY8uBXxjtP+lUWZBZwRsRJNpJKMHj7XoAfFio/+t6EFkFHpQmoV+Pi4B4o12eYZ/B4mItywVWGY8TxegByULelP+LKOvtxtlTxvlyUbf/VY0iYe3jncwn2yUudHP3LNGu9fRA9d5xme1HP3Mlxm+zjfae+VF9A/Q64Z/Fxm+vg58hn5OrgTu82sHcIP32RaRFRwg0iMGPwIfdfEmPQdA9MmRZcbrKvTANkBEqo0qduMSH8KFJwPdTCnVB/gp8JShr1FEyn2qzEIPFrvQX654pZQNPVisFpEP0b/2NcApRpvJwPPG62eBX4jI8/glbReRB4y2oL/c0DLSPRE9a7gKPbPwbtpux2cQFZHPMPImiz5Ng4i85dNnp6G/QGU+uvugZ4S5wGdGP65Ex2lOA541yr4Hfioir4nIevQXdzNQKSIfGH21Dv3l89pThZ5ZWYzP6W/oQbAJcKBneN56q9ED3uXAnejByykiHxvqqtGD1OdGvRSjXZHx3nD0F96O/gyOQM+o3jLK3kH/IGxDD6I2fWt5TfR0yI6e6WWKyG7jnnajXqPxY/wX43MAPQB769iNz+QC9ABuM8oKfeo40c9DFXrgsxt9sM/oy6noZ7MYPaiOAf5otH8WfXrqUMNXr/zA/me6HD1bP4r9z3i8ca/zgRuISY8Z/Np11EUpNRj4CbBEKWVVSq1AT/U/FJEl7Cdc8D1kKMAHSqnv0V/KfcC/lVLLlVJPKqUSfeqeAcwXkV3A/ehf1d3oB3yoUiodvYzyziBAz4a8g1qR8X8ouQBYaLzup5QqQH8ZvjJ01PvVPw8YrpR6WinVFz3zTVRKLVFKLVZKHWzUSwD2iMhG9NL9ZvQs6H7j9RrgJKMfp6JnbFmiT+YMRn8Zk33uG2e0XwIt+v8QYKFS6k6lVCF6ULpaKXUS+oejAv2l957/vlIptQ7947IeGIlehh8D3OZj/6noZ/c1w/77DP0T0T8Ka9Czrb3GlWnYuxu95fABMMB4LvKAxSLi/6zYgKeNshL0QFOLngVvR/8YLEQ/Qwq9PVGFHrxWoGeFnxtlqUC5j+6dhg0XAV8aup3oVUQmekXgAXLQ2wR72D/A7jT69ne0FO8zrdA/XgcZ/98AzEafpz/B0F3s1/ZOpdQPSqm/tUU00iulu9fdZi7gV2hKG+//56CXWIMx9vx83ktCf5lO8StPRS8brgD+YZTNYP+e3wDjbz/0ctMFHGKUPQj82XjtQD88Wezft8tE/3q/AfzLuP8S9LLv70a7cl97gTLj72D07GSKj62D0Q/86+iH2bfdregvwHj0DGkbeimahR7sVqNnS0+jB5ASQ8dU9N7OYKPsOkPfQ8BlRrvTgI/QX5SPDbteNuqX+/atj/1J6C/4DX79/6KP/d6yFwzblqCX29+jBwOv/SlG2QLD/rXG+6f42O/dh3vGx/6zjHb3+9j/AXoQ2ooeZDex/xk4yqc/d6CX7eN8+v9Z9I+tb1kaesC+26ifbuiqRQ9SCv3s7EY/m9XoH9FU9JL7e5/nsAQ9aL+G/oFINez7Er3E/Rw9c95l1N2E8ayiZ24lhq4i9BL5BPY/028b5S2ecaNfytGD+gz27/l5bXcaft/a3d/3LhtXutsAU0a2cdQFv8EPPQC9D1zbhp5b0V/qnehBo8h4eF/wq3c/UOrz/xHAO8brk4APjNenAk/51DvX56EbjP6SXmH8vx7NZLPaeODW+9TzH/yuN+xK8Knj/bLORg/MO9F7Wy70FzjbW8/n72Jgq4/ezYYNLiDXKKvwqa/Qy8jmfkTPvr417P/UKMsx/vfW24SeoXj/nw98jZ5h+uoaiF4K70XvZ5X62J/rU89rT7G3P33s/xQ92Pra722n0Etv38/yQUN/MfqLfys6X7QXKNgG3IM+ew5wG/pH7DZvmY+upwy/i4x2ZegZ2iafOt5B+0dgiM9zV/f/7Z1biF7VFcd/f4mFeCcavGBLUOoNUZRao7V2VNBaaSWiiEIfvGBi0ZaAL2IRtS+C1Ty0iEoiRUQRiViLNI4XxIkgjU0dyQUNOiYM5kGM1TEzE+u4fPiv43dmnIwxBMrMWT/44Hz77LP3Pufss85ae++1Tm4fkff2djw+2xx3f/aDdt/8Mn8T9Prqe1jj/yDTR/NaDWd7xvFQwk4m9/FmcqcpL4CRKefXRyoDXfjNFrP3G1cXST/AZuekCBQZKGEVsDkiHsi0hZIOy+35WHCsiIhjI2JRlvMKsFTSwZnvQDye8qGkE7P4i3CHBk+2PJnb24DFkg7I+i+iZ54fgzWZJ/L/c1iDBZunf5/uRDOA41Jga0Q0Y3iLWllOwg/neVhDHQbOxA9+wxL8QPTjMUIknYC11lOAXRExnHk/xCYceFxsC9baNmNT6o94ouZLYP+8tk37V2W+pqxV+EE7HY83juFvtjT35HI8hrgGeDgiFrTav6KVb0m2e5A00bL9R2HtZX1EDOc1bx7iB7L9Q80kk6QD8P16HgvNa3AfODmvDXndfoHv9614IuY6rB1ul3RGljUfTwK9GxFHZRkbsFC5rJXnkryG/wQuyLSr6YWLugYL763AYZJOyDy/xprqmdk3f5vnejHWEB/BffVF/HJfhF+uA+HJtrvyXvwGvwwOynNo+vhfgPsjYh7upxMRcbCko7PtexxoZM7w/5a+e/rDs3fv4jffHVgAbcdvuWFsjgQe+G2m7X+HJxTexjf1zill9mGT4Dj8oDVLGO7A40dv5rHPYhP3QGyGHNoq4278lt+AJzTWYm3kf/Q0tBuwOdO8fcfx+M4AvbfwRJYx0kr7Apu/Y63/zfKU7a06/kBvqUhk/vWZ56tM34q13tEsv2nXi63jmjBjzfZ41n9bpo1k+gietW3a9BW9JTu7Mm0MP/DN+Y7ldfnTlPv0BdaymnaPYZMxsLm4I9OGM20HfsG81WpXs1xkJ9bihrMdu7A5/J+8R6N5//rpac6Btd3VrfY3y20epLc8ZRxrfYO0+lOWOdLKM4hN2bez3HGsfW3OtJ3AY3hM7v3WOQ/hfnhf5t0GbMw+dlym7cTLdJZn+yey/JV5/d/DWvineU1fzzo+wy+0Q1r9vjF79zrQyGz/lYdHURSdZLaYvUVRFPuUEn5FUXSSEn5FUXSSEn5FUXSSEn5FUXSSEn5zAEkTkt6StEHS07m+bW/L+pukK3N75UzfSpXUJ+ncvajjA0nf+tDN7tKn5Pl8pv3T5L9L0m3ft43F3KeE39xgLByR41S8dm1Ze2cGXfjeRMSNEbFphix9eEF4Ucw6SvjNPQZwcIM+SQOSngM2pYP+fZLWpRP7UvDKfkl/lfSOpJewfyq571VJP8ntX0paL2lQ0ssZvGAZsDy1zp+nR83qrGOdpJ/lsYdL6pe0UdJKJnujTIukZyX9O4+5acq+FZn+sqSFmXa8pDV5zICkk/bFxSzmLrPtGx7FDKSGdyl2HwO7jZ0aEUMpQD6NiLMycsfrkvpxhJATsdvbkdiN79Ep5S7EARvOz7IWRMQOSQ8Bn0fEnzPfE9h9cK2kH2E3q5Oxt8HaiLhH0mXYs+S7uD7rmA+sk7Q6Ij7GXjZvRsRySXdm2bdg969lEbFF0tnYO6O+FFjslhJ+c4P5GS4JrPmtwubovyJiKNMvBk5rxvNw7L4f47iFT0bEBPZvfWWa8hfj+H5DABGxY5o8YF/SU+wmCsAhkg7KOq7IY5+X9Mlujm/ze0lLcvuH2daPsSvdU5n+OPBM1nEu8HSr7u6EZir2ihJ+c4OxcPTmb0gh0I7YK+DWiHhhSr5f7cN27AcsjohJcQZbAmmPkNSHBek5ETEq6VUcM3A6Iuv979RrUBQzUWN+3eEF4GZJ+4OjpGQEm9eAq3NM8GjggmmOfQM4X/6ANJIWZPoIk4Oa9uNgomS+Rhi9Rn5wWtKlOEjETByK4wWO5tjd4ta+/ehFx7kWm9Of4WguV2UdknT6d9RRdJwSft1hJR7PWy9pA/7ewzwccHRL7nsMx+GbRER8hL9v8YykQXpm5z+AJc2EB/k9k5xQ2URv1vluLDw3YvN3GzOzBpgnR3W+l144f7A2+9M8hwtxqHhwQNMbsn0bcfisotgtFdWlKIpOUppfURSdpIRfURSdpIRfURSdpIRfURSdpIRfURSdpIRfURSdpIRfURSdpIRfURSd5Gs6wT86W1x1NgAAAABJRU5ErkJggg==\n",
            "text/plain": [
              "<Figure size 432x288 with 2 Axes>"
            ]
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "Tensorboard\n",
        "-----------\n",
        "not working 👎"
      ],
      "metadata": {
        "id": "A6xT4uoraRqt"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "# %load_ext tensorboard"
      ],
      "metadata": {
        "id": "rLRe706BFkQS"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "# import datetime, os\n",
        "# logdir = os.path.join(\"logs\", datetime.datetime.now().strftime(\"%Y%m%d-%H%M%S\"))\n",
        "# tensorboard_callback = tf.keras.callbacks.TensorBoard(logdir, histogram_freq=1)"
      ],
      "metadata": {
        "id": "9rRYUpihF5h9"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "# %%time\n",
        "# history = model.fit(\n",
        "#     x=x_train,\n",
        "#     y=y_train,\n",
        "#     epochs=10,\n",
        "#     validation_data=(x_test, y_test),\n",
        "#     callbacks=[tensorboard_callback],\n",
        "#     verbose=1\n",
        "# )"
      ],
      "metadata": {
        "id": "HYSWcydfGO62"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "# # %tensorboard --logdir logs\n",
        "# from tensorboard import notebook\n",
        "# notebook.list() # View open TensorBoard instances\n",
        "\n",
        "# # Control TensorBoard display. If no port is provided, \n",
        "# # the most recently launched TensorBoard is used\n",
        "# notebook.display(port=6006, height=1000)"
      ],
      "metadata": {
        "id": "r8_KaTJHGh0z"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "**A better approach**\n",
        "Let's try with a bit more pre-processing\n",
        " - reduce MAXLEN to 512\n",
        " - remove most common words and least common words"
      ],
      "metadata": {
        "id": "zmnafwPPdHiU"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "from collections import defaultdict\n",
        "index_freq = defaultdict(int)\n",
        "max_index_in_wordmap = max(index_word.keys())\n",
        "for sent in x_train:\n",
        "  for index in sent:\n",
        "    # for some reason, a couple of indexes are not available in word_index map\n",
        "    if index > max_index_in_wordmap:\n",
        "      continue\n",
        "    else:\n",
        "      index_freq[index] += 1\n",
        "\n",
        "index_freq = [(index, freq) for index, freq in index_freq.items()]\n",
        "index_freq.sort(key=lambda x:x[1], reverse=True)\n",
        "index_freq[:2]"
      ],
      "metadata": {
        "id": "g90xt31cIhpS",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "d6363e3e-d710-4d67-9aed-9d8c808fa3f1"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "[(0, 7677835), (4, 65915)]"
            ]
          },
          "metadata": {},
          "execution_count": 61
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "word_freq = [(index_word[tup[0]], tup[1]) if tup[0] != 0 else (\"Unknown\", tup[1]) for tup in index_freq ]"
      ],
      "metadata": {
        "id": "7BeDoBlxeUUf"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "# word_freq[-25:]\n",
        "once_occuring = len([word for (word, count) in index_freq if count == 1])\n",
        "once_occuring"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "96ipIfS0h4m9",
        "outputId": "b78d198d-1d4d-41a9-e1c7-c5fe13e77365"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "9420"
            ]
          },
          "metadata": {},
          "execution_count": 73
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "(x_train, y_train), (x_test, y_test) = keras.datasets.reuters.load_data(skip_top=100, num_words=512)"
      ],
      "metadata": {
        "id": "9cePG-jhi_Hp"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "x_train.shape, x_test.shape, y_train.shape, y_test.shape"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "Pc27TMm-2qd6",
        "outputId": "05c43ce3-b69e-4ff6-a240-b68027d093d0"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "((8982,), (2246,), (8982,), (2246,))"
            ]
          },
          "metadata": {},
          "execution_count": 75
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "plot_lenghts(x_train)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 542
        },
        "id": "LYahB4OQ2xN-",
        "outputId": "fcc34e17-a65c-4772-fdec-08bb39e7b70e"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/html": [
              "<html>\n",
              "<head><meta charset=\"utf-8\" /></head>\n",
              "<body>\n",
              "    <div>            <script src=\"https://cdnjs.cloudflare.com/ajax/libs/mathjax/2.7.5/MathJax.js?config=TeX-AMS-MML_SVG\"></script><script type=\"text/javascript\">if (window.MathJax) {MathJax.Hub.Config({SVG: {font: \"STIX-Web\"}});}</script>                <script type=\"text/javascript\">window.PlotlyConfig = {MathJaxConfig: 'local'};</script>\n",
              "        <script src=\"https://cdn.plot.ly/plotly-2.8.3.min.js\"></script>                <div id=\"5aea0398-0681-4f84-a8c5-a90dbad8b1e4\" class=\"plotly-graph-div\" style=\"height:525px; width:100%;\"></div>            <script type=\"text/javascript\">                                    window.PLOTLYENV=window.PLOTLYENV || {};                                    if (document.getElementById(\"5aea0398-0681-4f84-a8c5-a90dbad8b1e4\")) {                    Plotly.newPlot(                        \"5aea0398-0681-4f84-a8c5-a90dbad8b1e4\",                        [{\"x\":[87,56,139,224,101,116,100,100,82,106,31,59,65,316,527,76,114,17,91,77,231,108,83,29,95,110,23,373,114,354,133,222,571,155,83,208,170,269,74,19,23,78,21,377,104,299,89,56,94,139,118,36,137,107,83,66,70,112,88,51,83,123,17,185,84,52,102,73,106,486,107,82,263,172,491,190,143,62,26,88,114,38,85,112,793,104,25,21,101,28,25,81,135,73,62,18,90,266,91,64,409,92,319,166,85,200,152,474,60,58,584,71,273,51,255,28,171,143,91,46,110,77,162,24,52,175,61,95,97,259,112,73,94,17,650,23,88,309,105,89,41,314,54,374,67,160,75,207,373,84,87,74,98,136,256,95,50,108,52,100,233,86,68,47,73,158,113,23,100,80,202,104,122,101,95,118,253,192,80,183,29,604,41,103,63,147,472,58,23,37,98,73,122,812,51,179,56,60,60,25,212,34,406,122,27,86,83,43,208,26,159,65,48,335,701,314,333,68,263,124,337,180,67,81,41,92,96,185,51,75,196,54,47,48,44,166,74,111,96,167,88,100,104,36,139,35,25,17,48,88,68,36,63,115,152,82,100,181,88,109,124,234,57,199,49,131,188,192,122,290,153,297,50,71,145,185,110,93,79,116,685,28,94,260,100,88,426,24,304,28,390,224,43,48,49,17,193,158,192,169,132,132,75,94,146,41,508,61,60,121,310,232,49,49,274,297,66,142,43,17,38,245,159,57,92,149,30,67,98,16,95,481,16,16,144,178,90,56,104,303,56,222,42,222,105,87,510,30,59,81,111,82,242,26,65,67,109,83,35,100,82,106,109,58,107,65,34,41,76,57,102,748,117,23,203,78,496,139,641,75,496,524,110,17,207,34,103,96,91,170,42,100,29,807,86,58,827,116,407,174,90,100,171,59,105,255,175,171,64,17,646,103,74,226,154,18,188,96,80,719,39,70,138,392,117,70,358,188,85,83,126,93,93,17,450,34,103,200,592,35,157,82,53,67,17,27,191,222,40,31,25,133,80,226,822,16,171,204,154,81,217,82,707,100,602,48,78,44,395,44,79,53,17,55,198,70,107,161,189,79,50,82,275,64,57,41,226,68,94,118,58,75,20,173,153,199,104,52,71,107,170,79,158,202,221,200,92,188,218,51,129,75,21,17,130,134,384,61,169,25,60,293,57,105,99,112,205,593,205,239,98,94,629,397,65,220,200,132,51,269,309,204,91,54,26,163,519,91,81,208,190,73,99,96,103,75,37,45,17,874,74,107,30,28,17,30,234,16,66,233,143,373,71,92,88,311,113,674,56,40,37,107,242,171,104,258,500,75,87,90,110,126,53,17,304,23,90,57,51,88,207,100,434,73,39,17,34,199,302,136,65,119,36,310,70,136,156,229,170,136,52,84,285,158,34,16,93,98,115,197,88,175,55,304,100,74,90,109,16,66,90,88,72,102,188,140,190,72,211,103,28,85,74,55,65,90,736,66,45,133,87,216,175,62,43,576,501,66,771,703,433,73,126,95,113,90,75,91,192,120,98,328,41,54,41,43,65,248,40,98,18,149,74,49,51,93,104,99,61,62,82,81,71,83,346,610,149,130,71,199,25,1032,87,368,143,57,59,166,217,505,32,37,70,643,24,58,49,62,287,52,693,17,722,282,192,24,58,169,89,123,564,269,109,44,173,229,184,97,583,100,41,47,58,377,68,86,17,50,177,102,600,73,192,67,79,53,78,56,187,17,17,115,94,79,67,238,85,313,200,17,95,130,143,63,244,267,18,183,178,191,78,86,40,97,232,93,83,25,90,91,36,40,150,93,204,68,78,27,316,154,383,110,80,105,43,66,105,110,60,81,160,17,56,53,149,83,177,144,63,92,244,90,39,109,455,80,17,99,189,113,74,130,64,708,105,250,68,43,33,92,90,140,746,17,91,65,146,256,485,102,17,55,51,89,58,92,93,100,225,85,113,143,86,52,113,302,101,136,296,143,88,102,171,156,251,412,143,290,21,310,63,158,198,90,77,374,17,34,186,90,50,56,90,120,130,45,96,108,185,44,187,59,17,67,269,114,234,57,94,734,102,93,240,402,108,82,17,50,65,18,77,49,65,186,153,26,113,363,49,295,81,96,137,279,308,453,29,101,138,91,253,229,84,899,126,92,192,49,184,49,19,177,241,171,22,112,472,79,49,101,41,25,69,41,142,232,52,221,87,44,35,16,306,108,41,102,629,27,254,95,65,261,25,287,416,94,63,27,133,51,626,17,442,25,321,99,46,34,101,75,809,311,743,115,228,287,171,92,57,43,53,30,98,287,166,98,91,17,81,165,140,99,194,89,115,117,48,162,484,197,17,224,89,140,73,115,78,252,23,42,104,92,93,99,72,129,197,261,193,484,137,63,96,47,19,170,85,57,303,153,113,79,130,81,94,48,103,84,43,94,24,194,125,97,61,206,17,32,768,25,90,94,59,61,65,363,32,291,30,175,290,183,128,56,274,76,52,30,96,281,429,84,266,252,84,92,55,98,32,518,60,249,70,573,55,58,107,320,53,331,87,245,122,95,392,92,431,60,651,206,36,94,516,444,149,405,102,333,49,216,631,17,118,149,19,69,42,53,114,103,86,96,470,138,39,96,86,126,171,112,168,73,89,43,141,60,665,119,202,546,36,85,55,78,57,377,83,17,62,111,86,110,89,288,52,65,72,103,83,40,240,403,87,110,333,21,249,69,17,62,98,203,51,375,38,491,75,99,76,319,187,48,78,80,159,293,205,70,188,25,102,56,87,281,134,280,82,302,229,59,199,182,23,63,121,282,58,752,418,76,52,28,162,110,81,100,159,26,242,408,116,56,139,162,29,217,89,113,36,167,87,90,249,160,35,72,35,58,97,23,91,191,105,102,65,84,367,97,51,108,48,237,621,81,89,369,320,163,103,164,98,94,86,30,422,279,48,233,35,145,78,246,79,86,91,16,148,84,77,261,65,80,32,808,29,196,17,58,201,43,192,79,127,113,242,42,46,184,60,82,99,103,87,47,158,176,42,97,30,81,225,40,223,186,152,504,95,100,88,108,174,109,164,94,59,193,128,97,297,210,169,27,78,17,33,89,102,249,88,62,33,144,200,174,92,214,97,103,15,192,295,89,79,45,52,210,121,329,52,115,152,88,136,17,97,227,61,329,95,260,190,241,99,40,55,167,401,196,154,321,63,183,67,73,67,206,151,72,75,74,52,61,141,93,101,67,224,31,59,657,43,526,110,114,536,105,280,215,85,80,27,78,68,57,105,176,17,17,74,86,41,136,91,107,82,245,51,82,62,79,62,174,391,65,262,26,63,172,75,153,165,266,27,55,49,59,266,185,68,42,103,267,69,57,17,551,289,46,111,103,88,87,249,134,596,50,226,43,130,52,304,63,212,17,52,91,165,73,114,122,204,28,203,102,193,95,76,413,43,67,58,109,161,68,147,35,17,116,59,324,104,59,248,68,94,193,17,80,130,371,44,148,110,74,76,393,79,17,482,96,294,57,58,144,21,81,137,108,80,188,24,84,219,108,170,80,48,176,47,19,274,433,26,81,372,68,26,83,109,92,154,192,17,93,74,67,96,27,85,178,634,44,163,77,129,65,89,69,164,32,533,199,97,255,195,132,101,39,117,115,62,128,102,312,71,141,93,356,35,167,90,305,324,53,143,467,160,383,310,438,282,381,204,85,24,111,90,41,51,42,108,282,244,127,904,48,816,57,108,83,99,193,78,222,301,85,40,659,50,242,27,48,144,88,123,163,339,199,386,416,43,89,35,49,211,82,21,248,306,164,99,279,612,601,74,383,287,568,102,64,65,139,81,91,231,126,82,70,153,69,165,103,354,89,202,186,422,94,220,92,241,145,70,62,68,61,191,116,54,16,17,95,89,38,180,174,41,106,215,55,38,23,82,172,226,16,53,229,69,96,42,82,27,66,299,187,284,55,33,84,118,200,46,492,28,159,249,66,190,159,150,115,696,95,83,121,154,77,17,91,198,213,53,284,51,269,754,75,165,459,144,262,522,213,65,49,80,17,38,157,292,147,342,52,336,54,82,76,100,71,27,16,593,91,105,37,17,133,25,67,158,47,186,175,71,774,409,206,172,77,66,386,89,163,252,78,75,67,111,101,217,77,181,108,98,42,36,61,262,67,607,285,17,430,169,107,79,51,71,95,109,257,50,42,138,91,54,90,17,261,236,111,82,99,88,275,42,37,51,79,132,32,75,52,265,53,236,165,50,335,103,56,17,52,141,107,156,93,280,342,93,79,214,51,28,696,302,69,105,71,17,60,215,584,49,103,86,618,37,17,636,54,17,194,17,45,111,146,183,94,316,75,73,82,314,28,64,483,288,303,35,156,261,362,245,199,122,369,553,17,104,45,266,72,99,234,50,97,35,206,117,492,17,280,67,81,66,17,83,102,101,37,52,58,17,75,25,75,196,354,138,134,693,196,94,91,121,77,41,26,99,86,91,87,74,17,72,98,497,471,34,29,206,50,21,76,54,50,26,87,24,417,100,24,672,225,67,352,280,143,121,88,49,167,84,131,219,47,96,90,71,48,542,57,23,217,98,88,84,88,77,53,254,2246,84,35,106,66,394,81,61,52,83,206,87,83,50,29,253,39,83,43,77,430,137,72,21,201,117,220,88,97,348,167,178,16,79,109,64,105,94,19,57,57,162,94,17,62,72,237,227,53,52,131,59,96,249,42,86,54,82,122,36,52,98,32,188,73,58,313,49,85,253,115,23,121,107,162,96,76,373,58,17,83,90,36,62,109,612,85,23,17,92,211,104,97,403,95,213,58,55,228,67,67,21,96,94,414,17,90,582,51,43,23,129,161,69,192,62,54,78,337,57,20,109,57,86,479,49,54,219,102,134,23,54,140,265,292,129,155,44,131,487,40,83,93,39,88,116,399,177,510,52,339,207,99,52,143,177,125,113,74,43,111,79,63,103,282,282,165,77,30,23,68,258,45,70,286,25,104,280,84,436,115,52,192,69,47,207,162,275,116,395,49,108,86,49,223,71,46,729,93,36,26,83,92,168,415,181,272,69,77,259,23,75,172,148,17,143,34,675,192,27,41,42,25,236,165,18,416,83,165,99,61,47,49,546,23,52,280,112,105,245,97,363,254,400,34,125,152,50,17,64,123,56,90,75,302,252,70,75,165,134,78,69,85,108,80,26,445,66,39,92,175,366,51,98,273,209,60,64,101,298,60,86,81,214,26,26,36,105,55,131,94,94,22,47,180,186,200,86,364,113,206,27,87,31,234,301,84,111,46,65,501,67,50,80,104,145,60,50,48,117,60,43,51,190,69,93,92,163,700,21,217,17,187,121,474,83,98,63,130,178,83,43,127,299,147,96,50,75,215,70,181,101,96,714,37,94,323,385,53,110,68,30,284,95,52,34,244,98,120,116,103,624,101,103,102,325,606,17,364,161,109,55,25,68,67,41,91,77,66,548,87,99,62,48,77,391,79,175,25,163,158,80,541,263,418,69,77,107,100,83,414,20,28,135,50,78,400,70,235,80,67,127,650,216,106,73,510,42,197,58,167,65,158,101,99,240,193,433,140,61,37,54,86,108,209,17,255,117,53,91,111,93,107,68,302,202,78,315,94,83,43,84,341,113,138,53,85,36,35,423,102,59,76,85,208,43,67,452,63,29,441,132,37,62,115,644,94,43,64,111,356,58,108,479,52,17,72,275,106,224,83,210,79,422,35,95,378,20,31,53,453,86,103,54,468,27,54,271,340,32,17,72,31,85,105,117,16,22,278,82,447,90,83,600,42,42,79,106,128,47,24,96,79,92,38,17,212,87,110,43,316,40,178,66,97,102,42,21,74,78,115,97,112,91,116,42,190,57,302,270,82,157,254,74,94,196,63,145,63,357,36,344,45,81,62,72,166,25,175,74,155,108,28,461,182,353,121,188,306,36,61,109,128,66,21,61,118,136,71,17,56,38,49,18,110,81,61,62,122,116,52,23,261,196,50,89,68,108,54,110,59,311,49,86,85,319,311,217,111,51,171,520,47,55,540,459,220,41,102,196,213,158,103,90,123,222,431,69,138,218,32,32,107,57,122,108,102,369,216,223,18,74,268,502,102,123,87,94,45,188,168,48,73,34,137,465,214,59,71,76,51,158,109,57,57,112,17,329,217,101,77,298,58,252,31,254,137,46,67,246,160,71,194,461,144,76,44,54,175,63,51,49,82,84,207,44,48,35,289,190,19,474,106,20,218,84,628,21,90,145,89,52,64,43,19,49,48,303,63,138,39,162,94,57,27,103,18,105,556,36,109,561,17,246,809,313,80,92,177,206,81,58,61,129,315,191,72,54,130,130,54,56,94,95,624,54,166,62,21,317,151,74,77,78,173,144,152,550,73,92,96,744,30,233,65,96,125,133,77,106,83,31,177,54,216,244,51,513,153,55,44,36,74,70,25,17,17,100,108,17,135,76,719,60,87,111,54,46,21,183,78,27,54,96,120,344,355,108,194,17,59,75,585,70,586,118,627,91,79,68,226,155,250,144,79,89,82,468,58,193,76,124,259,63,67,72,261,26,82,300,403,41,17,58,205,123,85,77,68,89,61,67,511,113,82,149,83,36,101,81,121,75,17,65,61,61,45,204,67,86,144,120,40,83,227,70,27,188,174,171,66,70,279,61,255,102,248,94,69,58,123,17,125,103,96,196,110,48,27,59,56,80,298,40,70,559,577,103,76,246,171,68,83,27,157,264,192,17,93,73,47,362,78,212,97,66,192,72,90,506,187,531,81,127,45,75,99,91,335,17,57,33,80,50,268,240,93,359,474,93,200,117,99,92,93,79,85,78,421,324,208,622,363,299,122,282,170,57,99,483,75,109,331,106,34,58,77,86,108,172,129,112,35,382,768,91,146,49,96,121,84,17,22,184,71,53,218,105,66,82,38,23,17,170,139,82,99,254,195,46,55,28,114,24,94,91,78,23,68,54,41,396,72,50,73,85,141,91,23,298,190,55,74,91,96,20,390,81,711,140,153,1053,310,74,65,391,61,37,49,79,36,25,17,476,90,197,49,39,97,92,35,342,101,248,197,50,117,175,208,22,504,67,63,51,32,62,60,41,163,67,97,375,32,57,84,38,381,66,94,107,93,19,98,104,84,107,41,67,43,741,108,55,38,113,67,76,38,75,278,330,469,59,108,76,92,148,270,96,41,18,224,54,67,94,277,184,101,23,266,186,64,214,162,55,92,17,133,100,91,66,105,122,63,291,144,595,113,181,35,57,246,148,267,135,210,73,84,176,193,103,74,86,72,350,85,109,75,249,72,27,322,43,63,90,124,121,34,172,541,96,94,117,278,67,53,53,106,395,113,55,393,23,77,53,97,154,95,17,66,59,187,53,307,149,45,29,42,99,40,48,48,229,544,84,278,20,160,94,85,199,298,76,65,58,104,41,93,67,152,63,23,31,828,76,42,485,105,118,66,227,18,96,93,35,76,61,81,181,219,280,53,41,274,204,163,223,98,38,40,115,104,320,455,102,45,120,52,94,172,370,385,178,109,514,178,346,58,128,93,47,244,82,68,35,128,116,483,66,49,191,101,240,329,88,72,106,176,25,63,111,208,125,216,60,82,87,40,73,123,35,28,58,135,172,223,319,17,102,99,504,303,51,424,17,137,27,85,62,47,76,49,105,109,71,40,49,42,76,98,94,231,50,59,83,63,307,130,99,113,213,80,175,591,195,174,54,93,94,59,49,140,651,98,83,178,160,62,334,91,68,41,533,628,33,99,65,140,93,21,112,98,158,345,95,134,31,114,158,73,213,547,109,194,89,108,472,75,65,85,245,37,234,23,91,77,17,122,70,47,100,57,93,108,67,363,378,91,102,119,29,109,359,270,280,70,50,105,67,76,59,48,68,147,47,218,17,55,24,76,78,97,103,67,71,40,68,235,267,44,17,142,35,89,91,87,227,216,83,108,476,109,25,85,503,90,424,107,36,61,99,78,100,81,68,35,171,642,210,101,38,358,17,37,87,78,445,84,295,358,676,73,55,165,115,106,246,113,438,283,196,195,117,475,96,170,83,27,260,162,31,35,21,80,64,38,100,86,162,98,96,60,81,77,55,79,41,121,55,94,67,100,99,105,134,76,648,778,204,165,231,309,83,91,96,74,442,116,63,98,100,84,414,75,172,50,81,50,34,474,331,87,64,167,304,71,110,65,53,60,401,59,82,48,99,25,184,224,37,66,87,99,89,36,85,27,106,65,85,92,241,88,54,209,111,218,17,194,40,225,95,785,183,48,16,88,92,278,37,65,21,61,175,27,78,59,74,50,79,47,61,84,69,49,272,705,881,103,111,117,73,47,347,76,55,97,59,43,331,302,218,101,17,74,187,71,57,158,87,25,74,630,54,35,94,101,67,67,109,243,210,153,28,143,296,17,24,30,250,173,606,115,119,69,68,69,86,45,46,84,293,60,76,196,30,95,205,114,92,68,670,111,36,59,191,77,96,104,39,209,94,25,58,17,212,475,314,356,27,158,110,17,111,113,105,60,42,85,106,18,27,69,240,103,128,287,193,59,92,232,67,54,27,51,25,34,155,57,293,50,79,179,98,72,74,69,87,302,192,101,72,74,211,245,73,56,590,97,59,112,102,64,134,83,29,524,193,27,62,102,370,93,106,213,88,67,16,123,103,44,537,17,61,156,27,318,197,81,276,481,76,476,89,86,102,96,184,91,427,93,41,212,84,119,37,658,97,252,59,99,188,274,97,25,176,86,39,167,90,35,84,292,74,105,97,81,217,28,129,34,42,130,129,163,144,90,17,96,114,217,51,104,105,166,483,27,391,106,194,185,671,61,25,100,21,139,175,55,120,251,84,207,191,170,147,267,209,169,53,211,650,208,588,475,159,46,62,224,43,23,110,258,43,125,86,116,472,698,26,85,187,27,86,163,56,44,92,66,259,119,336,57,57,165,25,320,205,16,113,26,212,241,227,90,52,91,154,62,187,62,186,35,100,34,102,88,276,23,25,99,77,19,22,208,76,130,25,116,520,25,199,91,115,115,121,71,82,476,289,243,61,306,138,164,250,70,109,118,230,40,71,59,684,83,291,16,141,92,71,353,139,194,184,112,62,237,51,191,17,34,288,304,76,723,27,193,74,146,118,87,116,267,514,26,80,174,31,25,28,78,223,165,79,72,58,41,92,66,235,121,63,116,366,81,571,88,72,45,95,17,129,58,35,43,130,395,94,21,104,848,67,62,460,113,225,19,419,226,44,107,83,69,36,62,349,99,85,59,85,59,73,17,27,104,619,108,219,97,38,57,248,61,72,211,55,526,301,115,91,135,153,47,67,309,68,91,59,44,118,143,17,212,184,67,43,162,184,93,39,53,90,17,273,26,450,234,593,170,51,121,312,94,78,186,69,407,51,66,55,609,53,60,287,66,96,140,163,44,96,90,250,82,1035,23,83,109,189,49,31,49,84,369,31,65,584,104,149,85,188,92,450,37,86,48,51,30,71,362,184,161,17,203,17,151,68,350,85,271,366,27,79,65,381,111,38,49,60,706,73,76,65,160,67,311,261,153,259,89,101,250,345,276,57,43,594,37,105,124,172,225,110,200,17,82,111,224,699,268,351,89,171,100,152,358,88,70,112,75,115,51,106,443,49,73,81,246,736,43,100,98,167,157,46,77,72,93,78,92,313,337,106,24,89,61,66,25,60,79,66,73,364,92,369,198,678,64,76,21,90,348,381,71,75,59,150,70,83,83,112,105,84,92,80,32,58,42,211,49,74,36,181,127,183,78,94,264,93,217,156,113,110,17,293,40,124,74,45,98,97,96,84,41,185,67,72,151,224,161,167,68,93,24,256,162,113,47,82,140,16,180,58,149,204,114,85,162,64,56,180,133,95,67,254,189,119,452,170,104,123,412,50,78,78,82,196,354,21,81,341,530,855,29,77,146,58,816,247,355,39,60,73,62,78,53,155,184,64,18,58,437,17,573,106,174,40,103,181,84,627,59,210,104,458,50,85,331,119,178,59,95,96,60,79,27,407,61,167,100,48,322,63,134,93,119,85,81,59,28,87,102,154,67,159,47,17,17,128,99,91,85,84,252,57,107,302,140,29,135,20,122,49,105,16,208,101,107,336,142,314,172,481,49,220,51,51,718,368,367,101,186,443,17,103,70,411,25,256,35,370,61,115,54,60,67,76,128,23,120,86,82,49,361,51,283,303,39,101,94,128,81,52,71,81,88,275,70,53,101,94,76,17,60,45,33,102,114,45,84,224,190,105,73,64,47,264,283,177,114,266,25,108,93,62,245,577,19,100,77,195,136,573,74,110,85,139,39,373,122,112,203,114,430,195,112,158,103,401,25,60,219,115,51,37,57,413,130,150,75,84,212,46,47,52,108,97,76,51,86,454,92,120,47,356,59,66,216,262,90,78,130,45,55,506,29,59,73,180,112,198,116,98,66,85,93,82,368,31,243,129,187,309,95,260,421,222,57,41,799,45,230,59,148,170,92,174,51,58,35,130,54,180,290,291,134,95,54,60,685,71,149,73,65,85,100,63,124,86,189,53,111,210,43,63,122,211,99,226,44,104,30,73,95,53,104,173,93,186,288,150,52,98,133,280,292,124,83,123,77,88,62,476,105,178,130,96,42,177,92,21,102,93,60,172,25,17,101,64,203,93,239,164,112,97,45,17,141,242,27,156,53,67,167,180,839,39,48,31,208,198,270,69,163,78,137,309,73,91,17,374,133,197,456,392,50,73,271,48,139,41,68,70,93,48,80,27,25,94,140,86,101,144,297,59,380,84,57,65,126,131,36,40,404,172,271,82,73,748,95,81,64,44,64,87,598,47,94,17,225,50,69,106,86,85,48,92,78,70,113,64,29,110,118,49,104,128,101,102,146,178,79,22,72,17,88,71,99,290,55,78,218,72,357,99,93,107,114,23,216,42,327,206,44,388,69,60,397,67,49,104,446,32,215,101,134,175,42,63,172,92,87,297,75,438,96,78,89,148,164,131,18,169,59,55,25,54,516,62,298,64,296,107,101,168,53,501,17,449,106,46,811,47,53,141,73,48,516,184,92,69,560,62,80,112,309,107,310,306,94,119,51,108,64,91,303,270,64,55,86,55,296,59,25,135,52,749,97,883,83,222,243,253,101,345,39,97,110,81,152,99,205,30,17,95,275,63,27,94,19,109,42,118,347,225,100,289,121,234,30,420,712,394,50,248,58,171,26,99,183,243,54,300,123,53,112,292,143,107,78,66,115,173,195,229,98,148,17,105,108,73,126,286,84,190,96,50,34,159,318,374,78,165,552,194,37,696,152,320,162,99,712,127,297,579,55,130,86,138,69,130,306,39,61,78,587,16,200,44,56,49,186,200,404,179,125,43,282,164,24,112,155,97,17,73,303,81,38,289,40,30,319,15,179,90,206,236,66,531,59,450,16,199,307,971,188,53,264,47,26,46,845,135,61,28,309,17,57,767,331,99,56,641,108,71,156,99,164,19,300,56,19,84,53,81,71,74,69,84,124,51,53,21,69,180,49,39,66,55,53,640,19,48,47,219,80,28,98,105,442,98,114,144,188,18,68,183,103,741,186,80,182,74,63,56,199,350,77,100,52,64,60,61,183,66,152,25,26,86,311,36,67,224,42,86,129,55,510,27,71,135,86,17,45,19,113,92,103,358,104,195,629,34,98,445,186,332,118,280,55,97,69,54,28,39,91,60,257,17,103,82,54,67,44,122,315,139,262,55,108,101,135,204,82,89,19,126,98,40,125,59,255,153,48,68,101,19,196,94,88,22,112,166,59,165,378,57,67,90,357,91,103,80,175,192,23,90,96,77,172,144,87,711,129,55,224,619,66,188,101,53,112,23,44,80,19,101,110,134,255,175,69,83,55,101,91,56,148,94,316,69,56,50,665,88,98,53,446,229,247,77,347,85,767,77,212,96,30,575,52,142,194,205,110,231,96,60,58,34,75,554,132,86,101,83,56,269,184,33,89,29,96,74,25,238,56,88,99,75,47,185,31,119,94,255,214,84,112,103,69,167,17,138,160,477,33,91,57,200,374,855,211,166,139,80,134,95,592,45,83,100,51,94,90,325,27,61,17,87,84,113,699,447,224,707,277,191,53,57,52,45,30,364,61,61,38,71,527,105,52,36,231,55,146,65,56,168,25,36,123,83,108,400,38,73,329,444,380,34,92,50,26,631,161,276,682,92,17,53,90,127,73,501,685,115,169,434,380,99,185,68,79,149,16,252,41,90,29,122,81,66,120,91,93,86,50,160,116,81,333,91,48,101,46,38,17,108,82,23,371,132,156,91,137,19,723,581,368,152,49,84,146,61,95,94,101,100,175,222,17,66,163,72,67,68,83,61,71,166,17,293,159,106,153,29,219,129,60,19,32,52,402,39,27,129,57,66,94,52,85,234,71,95,87,72,82,104,54,259,46,95,126,49,118,96,40,457,44,34,208,49,141,96,225,66,23,16,276,237,103,65,207,153,228,150,54,203,263,221,52,19,36,106,21,77,16,28,109,76,17,109,34,49,390,68,103,508,96,139,60,69,284,19,82,168,51,152,94,103,24,32,184,122,29,242,173,80,69,132,45,91,34,98,206,113,289,62,160,90,178,701,37,43,197,98,97,264,80,394,132,126,116,53,25,123,197,66,46,28,100,105,184,44,73,176,101,45,88,38,25,86,98,101,34,174,276,90,109,239,400,94,111,91,380,39,140,69,79,72,126,598,34,58,73,48,113,17,98,17,55,46,323,224,60,22,163,117,258,42,186,63,116,116,68,17,299,90,93,63,32,42,77,165,107,37,94,102,80,95,67,448,105,61,163,110,144,93,109,317,17,172,62,53,116,103,55,149,71,142,31,209,228,87,22,52,95,99,357,56,185,172,336,98,152,659,83,53,174,227,16,98,259,89,81,235,93,75,46,504,63,47,67,53,164,184,107,160,71,323,51,732,86,108,438,86,287,516,174,69,208,532,78,17,153,38,114,80,17,735,233,59,57,205,93,17,108,125,53,507,42,91,33,45,156,244,74,110,92,17,246,49,474,207,48,107,313,254,51,123,27,45,108,311,74,19,49,238,186,57,40,459,45,240,186,56,96,256,121,58,240,99,25,163,17,907,74,122,771,590,273,199,110,28,56,251,392,140,112,99,88,76,122,118,297,68,77,51,178,40,17,54,146,387,393,127,241,59,202,32,77,71,180,55,104,199,195,607,17,110,32,33,99,30,108,85,203,104,475,29,108,58,72,116,60,68,47,158,109,41,17,201,40,17,136,73,907,40,220,111,47,203,113,303,75,202,78,264,166,204,29,172,45,507,119,81,52,114,70,35,203,203,53,206,80,194,224,50,78,248,146,184,94,87,140,73,50,77,69,215,299,122,168,98,56,87,308,108,122,66,428,195,63,193,94,220,359,31,98,268,24,366,16,129,54,28,55,100,393,75,765,50,88,81,137,25,45,78,23,66,94,65,46,98,55,16,56,84,48,99,200,21,171,65,71,34,61,401,127,27,85,173,46,84,70,308,84,79,480,152,290,362,102,105,161,34,154,76,116,62,82,185,69,174,82,78,268,43,83,65,55,128,105,32,56,53,101,82,206,97,59,441,138,77,113,85,39,82,51,296,51,114,95,230,111,75,69,49,40,17,181,40,227,303,118,115,19,69,526,62,294,64,322,219,65,98,147,308,69,122,119,135,84,48,271,262,57,86,574,166,63,38,101,66,176,180,16,170,100,61,218,157,101,252,258,68,179,116,25,357,115,97,73,23,57,46,89,197,286,261,84,67,197,106,94,201,37,68,241,230,206,143,173,225,192,121,118,428,554,92,38,195,95,94,61,187,266,113,45,168,82,69,56,44,74,112,90,130,97,76,119,31,45,85,97,292,52,58,68,198,192,72,116,32,231,96,158,555,83,78,118,197,100,70,172,34,63,393,331,34,348,604,103,215,101,481,101,276,442,71,166,785,217,64,34,61,85,61,54,181,97,114,67,81,118,62,57,174,46,34,51,51,140,195,78,69,293,54,16,195,94,55,95,264,183,83,115,403,89,128,527,189,36,220,89,79,33,41,19,460,160,188,297,38,186,32,135,70,87,175,192,80,68,116,105,98,89,65,69,197,99,153,162,234,24,60,104,129,208,60,172,132,65,62,73,98,56,119,650,690,95,57,116,284,71,522,87,553,26,36,65,141,27,136,39,31,148,97,113,226,108,49,173,47,189,45,130,100,72,48,52,48,140,31,99,92,107,34,272,323,116,107,148,417,86,79,251,74,120,85,75,85,64,40,250,72,118,110,131,95,68,110,79,103,232,111,97,26,266,68,53,105,51,102,17,179,81,54,157,75,571,59,60,74,65,77,42,174,130,29,82,462,198,429,42,217,26,870,76,189,97,285,19,163,98,137,68,281,260,98,342,166,103,65,49,86,248,182,125,105,98,142,28,96,160,201,250,108,102,182,160,104,93,32,314,123,99,83,232,75,55,66,105,57,56,54,205,92,48,23,57,197,88,75,68,186,153,103,464,95,72,92,62,49,124,217,28,68,28,202,214,96,105,231,27,106,80,37,33,256,106,61,17,203,311,616,244,93,224,201,21,36,69,48,56,186,190,53,445,570,46,694,35,43,466,316,61,81,207,109,13,199,23,58,74,96,177,103,15,263,55,537,129,230,111,98,33,842,74,104,100,355,153,44,60,83,117,245,149,35,128,149,729,582,45,76,86,48,84,16,55,64,244,67,46,541,320,21,193,38,257,447,114,290,104,108,66,181,58,198,27,61,17,203,61,493,339,22,600,515,171,41,311,29,50,131,69,105,17,349,50,65,156,213,38,73,78,76,183,67,66,51,69,53,55,70,17,416,26,128,167,104,549,40,150,835,93,91,42,59,68,58,84,374,293,138,75,650,94,22,74,17,107,39,86,396,972,60,54,229,230,61,20,678,118,579,52,77,71,88,111,71,30,59,541,268,62,159,34,178,204,173,49,96,30,120,181,72,39,68,71,174,51,58,38,65,109,323,150,40,269,89,45,100,39,209,106,97,73,82,791,64,80,45,115,50,416,61,36,36,107,401,107,58,409,56,98,44,29,16,106,175,283,71,99,17,45,125,91,103,145,319,101,102,109,76,183,296,466,857,188,95,46,41,521,34,33,186,100,87,157,698,310,120,168,171,176,92,78,67,62,296,86,57,25,42,290,822,53,259,209,48,17,34,161,68,93,314,117,94,74,97,36,47,17,35,21,102,18,104,102,52,92,78,97,114,194,111,102,111,75,72,55,105,70,108,816,97,123,125,160,50,418,93,201,293,31,103,79,121,118,342,19,122,38,39,115,49,59,335,17,210,81,112,73,76,123,52,177,195,126,21,107,185,125,102,294,301,86,27,17,60,41,469,103,203,72,313,29,205,40,125,718,155,197,19,375,105,88,116,67,279,75,48,23,64,66,61,196,116,201,214,40,43,373,21,203,162,30,51,188,83,69,64,54,96,16,54,89,62,275,36,864,174,671,107,280,269,71,46,27,73,114,473,92,117,17,92,63,71,17,78,110,95,102,356,18,110,178,22,183,429,42,33,257,190,107,59,295,37,29,76,326,63,112,202,167,219,108,17,97,34,372,569,205,24,218,312,106,73,100,340,27,96,156,42,163,59,36,90,216,55,42,36,73,530,106,108,32,80,21,37,311,23,32,41,449,26,30,130,143,99,35,84,137,535,79,106,42,53,82,41,37,209,17,122,65,27,123,55,110,129,36,25,111,76,33,232,70,664,197,71,40,17,135,97,42,52,192,214,21,55,134,92,33,548,32,216,66,30,365,101,177,148,118,16,48,105,114,31,103,207,171,593,417,292,17,150,459,41,200,54,29,180,283,367,78,66,119,49,71,115,85,17,713,78,173,130,408,53,127,119,279,49,98,186,163,209,85,141,80,53,48,279,87,43,616,87,184,89,87,179,103,242,113,145,44,79,51,509,25,68,101,34,87,35,266,70,50,44,122,105,159,69,80,144,191,78,99,135,53,50,79,176,44,27,54,151,363,185,99,46,143,27,31,41,87,37,510,295,522,114,60,104,115,183,110,336,47,122,471,61,138,77,88,46,56,50,97,22,96,86,207,181,200,124,39,46,131,102,56,50,172,37,254,71,49,125,104,367,64,150,325,56,167,78,463,89,62,41,102,165,23,51,358,222,55,73,18,81,328,396,28,57,72,92,132,53,29,78,55,359,74,376,105,29,173,60,29,321,78,170,102,309,99,78,218,143,56,200,31,88,282,99,79,166,372,126,302,78,205,78,65,77,266,68,16,223,271,193,61,32,352,306,17,57,21,88,202,25,35,80,95,71,86,463,306,67,101,86,232,172,96,114,92,33,55,49,81,63,253,207,48,17,89,87,567,36,73,141,46,65,78,189,179,129,153,72,103,229,103,67,187,52,206,75,102,81,48,73,50,38,27,91,314,75,199,105,50,58,90,64,162,63,51,163,94,142,75,95,126,19,71,82,390,85,72,49,145,75,176,64,195,129,171,231,324,71,52,85,77,111,83,104,97,102,202,255,85,364,49,92,44,466,82,100,630,145,66,71,27,123,116,21,204,244,108,58,61,193,101,72,648,97,221,136,219,44,113,62,31,496,137,791,179,37,130,118,73,187,98,499,87,82,35,84,88,45,101,119,167,24,177,115,193,125,96,73,72,38,71,293,82,327,37,596,125,69,165,25,59,89,218,186,70,293,295,80,58,80,199,131,292,283,56,339,111,160,54,65,117,25,594,144,83,27,98,77,111,86,198,84,116,202,94,80,519,70,180,100,21,74,411,143,100,170,298,115,699,47,478,614,99,103,108,103,74,76,159,85,48,69,93,311,17,25,808,97,135,279,85,80,144,64,259,706,111,18,297,17,25,93,108,62,89,100,176,188,77,58,162,89,137,55,66,26,29,171,97,38,99,26,110,180,124,127,348,44,91,92,20,60,55,67,196,175,76,92,52,124,73,112,200,104,53,51,297,41,133,237,563,141,76,62,49,17,75,77,106,38,28,151,127,85,573,51,102,213,297,110,86,155,343,121,298,283,174,131,181,71,70,89,72,81,69,358,331,338,58,39,84,92,73,280,108,77,205,93,260,326,85,96,73,190,102,17,94,69,209,141,100,437,396,263,178,153,98,17,103,35,93,86,98,42,100,70,83,476,120,110,608,70,68,107,380,72,266,39,146,100,333,124,87,78,110,65,70,38,551,56,158,171,84,697,150,78,64,224,167,51,84,251,172,26,51,106,428,143,418,188,36,207,116,615,285,70,201,95,81,144,149,114,65,121,242,137,148,104,168,195,88,168,148,863,78,46,103,105,94,51,306,52,172,105,155,386,54,178,274,150,105,187,156,152,85,92,182,32,207,352,39,94,82,389,19,79,79,53,81,146,628,118,56,292,60,34,253,66,18,35,138,104,111,27,165,89,65,172,189,54,223,292,17,117,97,63,81,225,16,125,75,57,162,83,102,48,49,588,17,108,133,287,338,367,138,92,84,71,291,17,111,54,106,114,157,101,119,63,248,93,239,117,454,239,88,93,198,64,53,69,25,90,123,198,110,111,115,116,75,90,405,17,124,99,69,93,63,60,123,114,55,119,226,250,57,132,56,119,171,45,124,32,115,64,95,200,85,40,146,40,16,45,175,626,224,208,80,106,71,46,93,303,182,202,52,101,68,100,173,210,166,282,396,31,166,113,25,42,57,17,23,75,753,889,116,115,83,50,75,108,373,66,51,167,17,37,42,202,480,75,56,187,60,340,112,124,367,46,118,88,60,148,88,309,77,247,44,49,25,94,63,38,274,54,89,772,59,282,58,488,65,55,64,180,27,98,81,72,78,222,106,144,63,108,57,17,18,219,70,87,220,112,54,47,114,259,132,114,188,72,78,494,72,167,198,196,119,111,93,98,126,136,381,81,84,151,67,65,53,81,99,81,60,40,44,64,20,91,87,95,26,267,56,72,116,59,102,307,16,478,33,28,49,217,161,165,194,53,112,65,91,83,17,120,222,74,41,47,37,76,66,149,106,613,17,138,37,782,34,162,89,17,450,24,63,131,146,306,77,45,54,102,115,167,139,114,17,43,84,111,239,132,82,68,72,130,31,154,165,108,102,155,39,23,94,73,78,58,56,114,78,145,79,94,71,100,129,754,186,91,56,88,150,113,75,401,135,32,229,118,144,41,92,63,153,95,87,153,40,98,121,244,71,57,58,56,81,92,92,97,94,154,163,319,272,17,66,93,67,94,183,21,680,55,316,75,176,145,77,54,17,99,65,48,113,88,259,26,129,47,70,230,64,59,116,23,460,71,70,480,65,474,235,129,302,18,75,83,915,111,36,157,36,101,563,212,109,80,172,62,85,109,79,59,112,73,73,190,115,1332,27,25,108,73,346,207,205,92,129,190,46,157,61,247,648,211,53,74,199,73,92,131,53,177,70,39,98,17,145,111,59,41,428,99,91,75,390,282,123,19,36,139,98,59,92,434,89,517,122,62,83,225,32,204,70,78,47,107,100,92,39,90,23,451,63,294,211,252,90,104,56,185,71,71,88,26,57,101,46,17,53,17,86,44,102,447,73,151,203,129,24,92,41,180,93,103,59,186,82,102,272,50,60,93,92,39,180,301,220,21,57,119,113,73,17,148,143,343,365,778,488,166,41,17,89,21,17,46,186,849,81,192,17,86,192,30,189,148,48,269,89,52,31,299,112,83,81,117,88,44,77,186,130,89,334,545,87,44,92,129,372,109,87,101,48,52,82,93,41,222,63,561,34,80,17,327,251,86,204,25,27,235,193,301,85,72,694,29,243,45,77,109,342,73,42,72,17,35,48,114,467,113,42,103,100,82,99,30,102,224,753,111,102,30,510,56,94,99,83,631,83,187,82,33,178,77,172,62,317,63,29,114,53,78,73,137,308,60,130,97,357,97,85,101,121,86,49,70,228,48,718,490,49,139,123,33,101,19,81,176,183,48,262,45,39,108,160,95,51,16,85,68,49,49,26,100,81,85,243,165,28,77,16,21,72,21,51,19,184,383,240,466,58,367,59,539,18,58,41,106,466,350,231,231,65,287,86,67,35,160,96,73,17,49,48,357,401,218,464,60,595,114,66,66,553,69,103,78,173,45,59,124,37,158,308,183,111,79,98,48,141,298,69,224,93,276,80,64,56,190,106,25,192,59,126,23,99,45,94,57,27,86,32,89,119,94,60,77,54,26,287,60,17,99,86,244,101,50,105,17,277,117,26,171,97,206,134,152,82,42,238,338,55,57,149,18,173,89,243,130,29,282,313,170,214,218,119,159,253,89,76,201,289,283,461,346,668,52,69,107,20,178,101,78,100,53,537,117,205,171,103,87,170,36,17,83,152,68,49,192,186,204,97,481,70,67,66,208,83,236,22,74,108,40,94,93,183,105,31,56,52,381,46,150,117,63,54,286,52,83,178,28,162,145,163,51,61,168,95,58,17,75,81,126,230,116,27,17,173,95,63,47,458,46,31,195,104,70,68,28,313,82,173,85,35,101,17,221,268,83,2376,78,86,33,46,21,54,117,342,129,62,76,77,64,357,89,97,116,51,449,218,21,49,28,203,33,45,337,83,813,188,285,616,53,28,104,135,24,206,305,24,267,16,150,131,76,308,22,17,110,142,66,78,110,73,566,72,507,369,186,78,683,49,94,190,28,177,42,150,171,72,306,120,119,76,51,44,79,197,77,437,780,34,87,98,61,248,76,117,190,65,107,119,70,81,41,169,19,73,46,49,51,21,31,91,95,81,285,24,36,78,100,233,65,150,93,85,73,31,47,62,17,48,74,129,118,94,61,54,33,688,80,57,150,90,53,230,186,79,45,225,111,90,504,78,825,91,116,39,224,184,428,87,66,260,205,75,54,19,48,108,99,118,41,80,234,78,468,97,59,90,233,315,19,84,21,175,111,95,91,187,69,104,122,99,83,149,73,87,51,215,123,266,45,73,165,170,120,151,70,95,64,618,80,19,35,186,69,82,48,69,58,189,74,530,88,335,152,146,457,42,269,129,27,68,116,133,148,163,195,176,17,149,48,250,59,18,54,44,211,88,62,314,249,92,43,53,72,101,17,250,49,76,92,102,30,86,36,109,141,62,437,588,19,252,58,64,108,116,257,120,189,285,88,82,70,30,145,197,214,77,119,105],\"type\":\"histogram\"}],                        {\"template\":{\"data\":{\"bar\":[{\"error_x\":{\"color\":\"#2a3f5f\"},\"error_y\":{\"color\":\"#2a3f5f\"},\"marker\":{\"line\":{\"color\":\"#E5ECF6\",\"width\":0.5},\"pattern\":{\"fillmode\":\"overlay\",\"size\":10,\"solidity\":0.2}},\"type\":\"bar\"}],\"barpolar\":[{\"marker\":{\"line\":{\"color\":\"#E5ECF6\",\"width\":0.5},\"pattern\":{\"fillmode\":\"overlay\",\"size\":10,\"solidity\":0.2}},\"type\":\"barpolar\"}],\"carpet\":[{\"aaxis\":{\"endlinecolor\":\"#2a3f5f\",\"gridcolor\":\"white\",\"linecolor\":\"white\",\"minorgridcolor\":\"white\",\"startlinecolor\":\"#2a3f5f\"},\"baxis\":{\"endlinecolor\":\"#2a3f5f\",\"gridcolor\":\"white\",\"linecolor\":\"white\",\"minorgridcolor\":\"white\",\"startlinecolor\":\"#2a3f5f\"},\"type\":\"carpet\"}],\"choropleth\":[{\"colorbar\":{\"outlinewidth\":0,\"ticks\":\"\"},\"type\":\"choropleth\"}],\"contour\":[{\"colorbar\":{\"outlinewidth\":0,\"ticks\":\"\"},\"colorscale\":[[0.0,\"#0d0887\"],[0.1111111111111111,\"#46039f\"],[0.2222222222222222,\"#7201a8\"],[0.3333333333333333,\"#9c179e\"],[0.4444444444444444,\"#bd3786\"],[0.5555555555555556,\"#d8576b\"],[0.6666666666666666,\"#ed7953\"],[0.7777777777777778,\"#fb9f3a\"],[0.8888888888888888,\"#fdca26\"],[1.0,\"#f0f921\"]],\"type\":\"contour\"}],\"contourcarpet\":[{\"colorbar\":{\"outlinewidth\":0,\"ticks\":\"\"},\"type\":\"contourcarpet\"}],\"heatmap\":[{\"colorbar\":{\"outlinewidth\":0,\"ticks\":\"\"},\"colorscale\":[[0.0,\"#0d0887\"],[0.1111111111111111,\"#46039f\"],[0.2222222222222222,\"#7201a8\"],[0.3333333333333333,\"#9c179e\"],[0.4444444444444444,\"#bd3786\"],[0.5555555555555556,\"#d8576b\"],[0.6666666666666666,\"#ed7953\"],[0.7777777777777778,\"#fb9f3a\"],[0.8888888888888888,\"#fdca26\"],[1.0,\"#f0f921\"]],\"type\":\"heatmap\"}],\"heatmapgl\":[{\"colorbar\":{\"outlinewidth\":0,\"ticks\":\"\"},\"colorscale\":[[0.0,\"#0d0887\"],[0.1111111111111111,\"#46039f\"],[0.2222222222222222,\"#7201a8\"],[0.3333333333333333,\"#9c179e\"],[0.4444444444444444,\"#bd3786\"],[0.5555555555555556,\"#d8576b\"],[0.6666666666666666,\"#ed7953\"],[0.7777777777777778,\"#fb9f3a\"],[0.8888888888888888,\"#fdca26\"],[1.0,\"#f0f921\"]],\"type\":\"heatmapgl\"}],\"histogram\":[{\"marker\":{\"pattern\":{\"fillmode\":\"overlay\",\"size\":10,\"solidity\":0.2}},\"type\":\"histogram\"}],\"histogram2d\":[{\"colorbar\":{\"outlinewidth\":0,\"ticks\":\"\"},\"colorscale\":[[0.0,\"#0d0887\"],[0.1111111111111111,\"#46039f\"],[0.2222222222222222,\"#7201a8\"],[0.3333333333333333,\"#9c179e\"],[0.4444444444444444,\"#bd3786\"],[0.5555555555555556,\"#d8576b\"],[0.6666666666666666,\"#ed7953\"],[0.7777777777777778,\"#fb9f3a\"],[0.8888888888888888,\"#fdca26\"],[1.0,\"#f0f921\"]],\"type\":\"histogram2d\"}],\"histogram2dcontour\":[{\"colorbar\":{\"outlinewidth\":0,\"ticks\":\"\"},\"colorscale\":[[0.0,\"#0d0887\"],[0.1111111111111111,\"#46039f\"],[0.2222222222222222,\"#7201a8\"],[0.3333333333333333,\"#9c179e\"],[0.4444444444444444,\"#bd3786\"],[0.5555555555555556,\"#d8576b\"],[0.6666666666666666,\"#ed7953\"],[0.7777777777777778,\"#fb9f3a\"],[0.8888888888888888,\"#fdca26\"],[1.0,\"#f0f921\"]],\"type\":\"histogram2dcontour\"}],\"mesh3d\":[{\"colorbar\":{\"outlinewidth\":0,\"ticks\":\"\"},\"type\":\"mesh3d\"}],\"parcoords\":[{\"line\":{\"colorbar\":{\"outlinewidth\":0,\"ticks\":\"\"}},\"type\":\"parcoords\"}],\"pie\":[{\"automargin\":true,\"type\":\"pie\"}],\"scatter\":[{\"marker\":{\"colorbar\":{\"outlinewidth\":0,\"ticks\":\"\"}},\"type\":\"scatter\"}],\"scatter3d\":[{\"line\":{\"colorbar\":{\"outlinewidth\":0,\"ticks\":\"\"}},\"marker\":{\"colorbar\":{\"outlinewidth\":0,\"ticks\":\"\"}},\"type\":\"scatter3d\"}],\"scattercarpet\":[{\"marker\":{\"colorbar\":{\"outlinewidth\":0,\"ticks\":\"\"}},\"type\":\"scattercarpet\"}],\"scattergeo\":[{\"marker\":{\"colorbar\":{\"outlinewidth\":0,\"ticks\":\"\"}},\"type\":\"scattergeo\"}],\"scattergl\":[{\"marker\":{\"colorbar\":{\"outlinewidth\":0,\"ticks\":\"\"}},\"type\":\"scattergl\"}],\"scattermapbox\":[{\"marker\":{\"colorbar\":{\"outlinewidth\":0,\"ticks\":\"\"}},\"type\":\"scattermapbox\"}],\"scatterpolar\":[{\"marker\":{\"colorbar\":{\"outlinewidth\":0,\"ticks\":\"\"}},\"type\":\"scatterpolar\"}],\"scatterpolargl\":[{\"marker\":{\"colorbar\":{\"outlinewidth\":0,\"ticks\":\"\"}},\"type\":\"scatterpolargl\"}],\"scatterternary\":[{\"marker\":{\"colorbar\":{\"outlinewidth\":0,\"ticks\":\"\"}},\"type\":\"scatterternary\"}],\"surface\":[{\"colorbar\":{\"outlinewidth\":0,\"ticks\":\"\"},\"colorscale\":[[0.0,\"#0d0887\"],[0.1111111111111111,\"#46039f\"],[0.2222222222222222,\"#7201a8\"],[0.3333333333333333,\"#9c179e\"],[0.4444444444444444,\"#bd3786\"],[0.5555555555555556,\"#d8576b\"],[0.6666666666666666,\"#ed7953\"],[0.7777777777777778,\"#fb9f3a\"],[0.8888888888888888,\"#fdca26\"],[1.0,\"#f0f921\"]],\"type\":\"surface\"}],\"table\":[{\"cells\":{\"fill\":{\"color\":\"#EBF0F8\"},\"line\":{\"color\":\"white\"}},\"header\":{\"fill\":{\"color\":\"#C8D4E3\"},\"line\":{\"color\":\"white\"}},\"type\":\"table\"}]},\"layout\":{\"annotationdefaults\":{\"arrowcolor\":\"#2a3f5f\",\"arrowhead\":0,\"arrowwidth\":1},\"autotypenumbers\":\"strict\",\"coloraxis\":{\"colorbar\":{\"outlinewidth\":0,\"ticks\":\"\"}},\"colorscale\":{\"diverging\":[[0,\"#8e0152\"],[0.1,\"#c51b7d\"],[0.2,\"#de77ae\"],[0.3,\"#f1b6da\"],[0.4,\"#fde0ef\"],[0.5,\"#f7f7f7\"],[0.6,\"#e6f5d0\"],[0.7,\"#b8e186\"],[0.8,\"#7fbc41\"],[0.9,\"#4d9221\"],[1,\"#276419\"]],\"sequential\":[[0.0,\"#0d0887\"],[0.1111111111111111,\"#46039f\"],[0.2222222222222222,\"#7201a8\"],[0.3333333333333333,\"#9c179e\"],[0.4444444444444444,\"#bd3786\"],[0.5555555555555556,\"#d8576b\"],[0.6666666666666666,\"#ed7953\"],[0.7777777777777778,\"#fb9f3a\"],[0.8888888888888888,\"#fdca26\"],[1.0,\"#f0f921\"]],\"sequentialminus\":[[0.0,\"#0d0887\"],[0.1111111111111111,\"#46039f\"],[0.2222222222222222,\"#7201a8\"],[0.3333333333333333,\"#9c179e\"],[0.4444444444444444,\"#bd3786\"],[0.5555555555555556,\"#d8576b\"],[0.6666666666666666,\"#ed7953\"],[0.7777777777777778,\"#fb9f3a\"],[0.8888888888888888,\"#fdca26\"],[1.0,\"#f0f921\"]]},\"colorway\":[\"#636efa\",\"#EF553B\",\"#00cc96\",\"#ab63fa\",\"#FFA15A\",\"#19d3f3\",\"#FF6692\",\"#B6E880\",\"#FF97FF\",\"#FECB52\"],\"font\":{\"color\":\"#2a3f5f\"},\"geo\":{\"bgcolor\":\"white\",\"lakecolor\":\"white\",\"landcolor\":\"#E5ECF6\",\"showlakes\":true,\"showland\":true,\"subunitcolor\":\"white\"},\"hoverlabel\":{\"align\":\"left\"},\"hovermode\":\"closest\",\"mapbox\":{\"style\":\"light\"},\"paper_bgcolor\":\"white\",\"plot_bgcolor\":\"#E5ECF6\",\"polar\":{\"angularaxis\":{\"gridcolor\":\"white\",\"linecolor\":\"white\",\"ticks\":\"\"},\"bgcolor\":\"#E5ECF6\",\"radialaxis\":{\"gridcolor\":\"white\",\"linecolor\":\"white\",\"ticks\":\"\"}},\"scene\":{\"xaxis\":{\"backgroundcolor\":\"#E5ECF6\",\"gridcolor\":\"white\",\"gridwidth\":2,\"linecolor\":\"white\",\"showbackground\":true,\"ticks\":\"\",\"zerolinecolor\":\"white\"},\"yaxis\":{\"backgroundcolor\":\"#E5ECF6\",\"gridcolor\":\"white\",\"gridwidth\":2,\"linecolor\":\"white\",\"showbackground\":true,\"ticks\":\"\",\"zerolinecolor\":\"white\"},\"zaxis\":{\"backgroundcolor\":\"#E5ECF6\",\"gridcolor\":\"white\",\"gridwidth\":2,\"linecolor\":\"white\",\"showbackground\":true,\"ticks\":\"\",\"zerolinecolor\":\"white\"}},\"shapedefaults\":{\"line\":{\"color\":\"#2a3f5f\"}},\"ternary\":{\"aaxis\":{\"gridcolor\":\"white\",\"linecolor\":\"white\",\"ticks\":\"\"},\"baxis\":{\"gridcolor\":\"white\",\"linecolor\":\"white\",\"ticks\":\"\"},\"bgcolor\":\"#E5ECF6\",\"caxis\":{\"gridcolor\":\"white\",\"linecolor\":\"white\",\"ticks\":\"\"}},\"title\":{\"x\":0.05},\"xaxis\":{\"automargin\":true,\"gridcolor\":\"white\",\"linecolor\":\"white\",\"ticks\":\"\",\"title\":{\"standoff\":15},\"zerolinecolor\":\"white\",\"zerolinewidth\":2},\"yaxis\":{\"automargin\":true,\"gridcolor\":\"white\",\"linecolor\":\"white\",\"ticks\":\"\",\"title\":{\"standoff\":15},\"zerolinecolor\":\"white\",\"zerolinewidth\":2}}}},                        {\"responsive\": true}                    ).then(function(){\n",
              "                            \n",
              "var gd = document.getElementById('5aea0398-0681-4f84-a8c5-a90dbad8b1e4');\n",
              "var x = new MutationObserver(function (mutations, observer) {{\n",
              "        var display = window.getComputedStyle(gd).display;\n",
              "        if (!display || display === 'none') {{\n",
              "            console.log([gd, 'removed!']);\n",
              "            Plotly.purge(gd);\n",
              "            observer.disconnect();\n",
              "        }}\n",
              "}});\n",
              "\n",
              "// Listen for the removal of the full notebook cells\n",
              "var notebookContainer = gd.closest('#notebook-container');\n",
              "if (notebookContainer) {{\n",
              "    x.observe(notebookContainer, {childList: true});\n",
              "}}\n",
              "\n",
              "// Listen for the clearing of the current output cell\n",
              "var outputEl = gd.closest('.output');\n",
              "if (outputEl) {{\n",
              "    x.observe(outputEl, {childList: true});\n",
              "}}\n",
              "\n",
              "                        })                };                            </script>        </div>\n",
              "</body>\n",
              "</html>"
            ]
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        ""
      ],
      "metadata": {
        "id": "pViXDInD3dhU"
      },
      "execution_count": null,
      "outputs": []
    }
  ]
}