Just_Bert.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "name": "Just_Bert.ipynb",
      "provenance": [],
      "collapsed_sections": [],
      "machine_shape": "hm",
      "include_colab_link": true
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    },
    "accelerator": "GPU"
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "view-in-github",
        "colab_type": "text"
      },
      "source": [
        "<a href=\"https://colab.research.google.com/github/AyHaski/DataScienceSS20Project/blob/master/Just_Bert.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "2sCDfe2eQFLI",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 122
        },
        "outputId": "8e97a123-f4c3-4354-be29-cca569150ddc"
      },
      "source": [
        "from google.colab import drive\n",
        "drive.mount('/content/drive')"
      ],
      "execution_count": 2,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Go to this URL in a browser: https://accounts.google.com/o/oauth2/auth?client_id=947318989803-6bn6qk8qdgf4n4g3pfee6491hc0brc4i.apps.googleusercontent.com&redirect_uri=urn%3aietf%3awg%3aoauth%3a2.0%3aoob&response_type=code&scope=email%20https%3a%2f%2fwww.googleapis.com%2fauth%2fdocs.test%20https%3a%2f%2fwww.googleapis.com%2fauth%2fdrive%20https%3a%2f%2fwww.googleapis.com%2fauth%2fdrive.photos.readonly%20https%3a%2f%2fwww.googleapis.com%2fauth%2fpeopleapi.readonly\n",
            "\n",
            "Enter your authorization code:\n",
            "··········\n",
            "Mounted at /content/drive\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "L47xtsSDUf2E",
        "colab_type": "code",
        "outputId": "fdb1013c-5fcc-4c88-a2cd-c8b2c94fd7db",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1000
        }
      },
      "source": [
        "!git clone https://github.com/AyHaski/DataScienceSS20Project\n",
        "!pip install tensorflow==2.0\n",
        "!pip install tensorflow_hub\n",
        "!pip install bert-for-tf2\n",
        "!pip install sentencepiece"
      ],
      "execution_count": 21,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "fatal: destination path 'DataScienceSS20Project' already exists and is not an empty directory.\n",
            "Collecting tensorflow==2.0\n",
            "  Using cached https://files.pythonhosted.org/packages/46/0f/7bd55361168bb32796b360ad15a25de6966c9c1beb58a8e30c01c8279862/tensorflow-2.0.0-cp36-cp36m-manylinux2010_x86_64.whl\n",
            "Requirement already satisfied: google-pasta>=0.1.6 in /usr/local/lib/python3.6/dist-packages (from tensorflow==2.0) (0.2.0)\n",
            "Requirement already satisfied: keras-applications>=1.0.8 in /usr/local/lib/python3.6/dist-packages (from tensorflow==2.0) (1.0.8)\n",
            "Collecting gast==0.2.2\n",
            "  Using cached https://files.pythonhosted.org/packages/4e/35/11749bf99b2d4e3cceb4d55ca22590b0d7c2c62b9de38ac4a4a7f4687421/gast-0.2.2.tar.gz\n",
            "Requirement already satisfied: absl-py>=0.7.0 in /usr/local/lib/python3.6/dist-packages (from tensorflow==2.0) (0.9.0)\n",
            "Requirement already satisfied: protobuf>=3.6.1 in /usr/local/lib/python3.6/dist-packages (from tensorflow==2.0) (3.10.0)\n",
            "Requirement already satisfied: grpcio>=1.8.6 in /usr/local/lib/python3.6/dist-packages (from tensorflow==2.0) (1.29.0)\n",
            "Requirement already satisfied: wheel>=0.26 in /usr/local/lib/python3.6/dist-packages (from tensorflow==2.0) (0.34.2)\n",
            "Requirement already satisfied: termcolor>=1.1.0 in /usr/local/lib/python3.6/dist-packages (from tensorflow==2.0) (1.1.0)\n",
            "Requirement already satisfied: opt-einsum>=2.3.2 in /usr/local/lib/python3.6/dist-packages (from tensorflow==2.0) (3.2.1)\n",
            "Requirement already satisfied: wrapt>=1.11.1 in /usr/local/lib/python3.6/dist-packages (from tensorflow==2.0) (1.12.1)\n",
            "Requirement already satisfied: keras-preprocessing>=1.0.5 in /usr/local/lib/python3.6/dist-packages (from tensorflow==2.0) (1.1.2)\n",
            "Collecting tensorboard<2.1.0,>=2.0.0\n",
            "  Using cached https://files.pythonhosted.org/packages/76/54/99b9d5d52d5cb732f099baaaf7740403e83fe6b0cedde940fabd2b13d75a/tensorboard-2.0.2-py3-none-any.whl\n",
            "Collecting tensorflow-estimator<2.1.0,>=2.0.0\n",
            "  Using cached https://files.pythonhosted.org/packages/fc/08/8b927337b7019c374719145d1dceba21a8bb909b93b1ad6f8fb7d22c1ca1/tensorflow_estimator-2.0.1-py2.py3-none-any.whl\n",
            "Requirement already satisfied: six>=1.10.0 in /usr/local/lib/python3.6/dist-packages (from tensorflow==2.0) (1.12.0)\n",
            "Requirement already satisfied: numpy<2.0,>=1.16.0 in /usr/local/lib/python3.6/dist-packages (from tensorflow==2.0) (1.18.5)\n",
            "Requirement already satisfied: astor>=0.6.0 in /usr/local/lib/python3.6/dist-packages (from tensorflow==2.0) (0.8.1)\n",
            "Requirement already satisfied: h5py in /usr/local/lib/python3.6/dist-packages (from keras-applications>=1.0.8->tensorflow==2.0) (2.10.0)\n",
            "Requirement already satisfied: setuptools in /usr/local/lib/python3.6/dist-packages (from protobuf>=3.6.1->tensorflow==2.0) (47.1.1)\n",
            "Requirement already satisfied: werkzeug>=0.11.15 in /usr/local/lib/python3.6/dist-packages (from tensorboard<2.1.0,>=2.0.0->tensorflow==2.0) (1.0.1)\n",
            "Requirement already satisfied: google-auth<2,>=1.6.3 in /usr/local/lib/python3.6/dist-packages (from tensorboard<2.1.0,>=2.0.0->tensorflow==2.0) (1.7.2)\n",
            "Requirement already satisfied: markdown>=2.6.8 in /usr/local/lib/python3.6/dist-packages (from tensorboard<2.1.0,>=2.0.0->tensorflow==2.0) (3.2.2)\n",
            "Requirement already satisfied: requests<3,>=2.21.0 in /usr/local/lib/python3.6/dist-packages (from tensorboard<2.1.0,>=2.0.0->tensorflow==2.0) (2.23.0)\n",
            "Requirement already satisfied: google-auth-oauthlib<0.5,>=0.4.1 in /usr/local/lib/python3.6/dist-packages (from tensorboard<2.1.0,>=2.0.0->tensorflow==2.0) (0.4.1)\n",
            "Requirement already satisfied: cachetools<3.2,>=2.0.0 in /usr/local/lib/python3.6/dist-packages (from google-auth<2,>=1.6.3->tensorboard<2.1.0,>=2.0.0->tensorflow==2.0) (3.1.1)\n",
            "Requirement already satisfied: rsa<4.1,>=3.1.4 in /usr/local/lib/python3.6/dist-packages (from google-auth<2,>=1.6.3->tensorboard<2.1.0,>=2.0.0->tensorflow==2.0) (4.0)\n",
            "Requirement already satisfied: pyasn1-modules>=0.2.1 in /usr/local/lib/python3.6/dist-packages (from google-auth<2,>=1.6.3->tensorboard<2.1.0,>=2.0.0->tensorflow==2.0) (0.2.8)\n",
            "Requirement already satisfied: importlib-metadata; python_version < \"3.8\" in /usr/local/lib/python3.6/dist-packages (from markdown>=2.6.8->tensorboard<2.1.0,>=2.0.0->tensorflow==2.0) (1.6.0)\n",
            "Requirement already satisfied: idna<3,>=2.5 in /usr/local/lib/python3.6/dist-packages (from requests<3,>=2.21.0->tensorboard<2.1.0,>=2.0.0->tensorflow==2.0) (2.9)\n",
            "Requirement already satisfied: chardet<4,>=3.0.2 in /usr/local/lib/python3.6/dist-packages (from requests<3,>=2.21.0->tensorboard<2.1.0,>=2.0.0->tensorflow==2.0) (3.0.4)\n",
            "Requirement already satisfied: urllib3!=1.25.0,!=1.25.1,<1.26,>=1.21.1 in /usr/local/lib/python3.6/dist-packages (from requests<3,>=2.21.0->tensorboard<2.1.0,>=2.0.0->tensorflow==2.0) (1.24.3)\n",
            "Requirement already satisfied: certifi>=2017.4.17 in /usr/local/lib/python3.6/dist-packages (from requests<3,>=2.21.0->tensorboard<2.1.0,>=2.0.0->tensorflow==2.0) (2020.4.5.1)\n",
            "Requirement already satisfied: requests-oauthlib>=0.7.0 in /usr/local/lib/python3.6/dist-packages (from google-auth-oauthlib<0.5,>=0.4.1->tensorboard<2.1.0,>=2.0.0->tensorflow==2.0) (1.3.0)\n",
            "Requirement already satisfied: pyasn1>=0.1.3 in /usr/local/lib/python3.6/dist-packages (from rsa<4.1,>=3.1.4->google-auth<2,>=1.6.3->tensorboard<2.1.0,>=2.0.0->tensorflow==2.0) (0.4.8)\n",
            "Requirement already satisfied: zipp>=0.5 in /usr/local/lib/python3.6/dist-packages (from importlib-metadata; python_version < \"3.8\"->markdown>=2.6.8->tensorboard<2.1.0,>=2.0.0->tensorflow==2.0) (3.1.0)\n",
            "Requirement already satisfied: oauthlib>=3.0.0 in /usr/local/lib/python3.6/dist-packages (from requests-oauthlib>=0.7.0->google-auth-oauthlib<0.5,>=0.4.1->tensorboard<2.1.0,>=2.0.0->tensorflow==2.0) (3.1.0)\n",
            "Building wheels for collected packages: gast\n",
            "  Building wheel for gast (setup.py) ... \u001b[?25l\u001b[?25hdone\n",
            "  Created wheel for gast: filename=gast-0.2.2-cp36-none-any.whl size=7540 sha256=da1468d3de58932aa1cab24ac2097e2b48fcad2b71c43600158a63a94b40163e\n",
            "  Stored in directory: /root/.cache/pip/wheels/5c/2e/7e/a1d4d4fcebe6c381f378ce7743a3ced3699feb89bcfbdadadd\n",
            "Successfully built gast\n",
            "\u001b[31mERROR: tensorflow-probability 0.10.0 has requirement gast>=0.3.2, but you'll have gast 0.2.2 which is incompatible.\u001b[0m\n",
            "Installing collected packages: gast, tensorboard, tensorflow-estimator, tensorflow\n",
            "  Found existing installation: gast 0.3.3\n",
            "    Uninstalling gast-0.3.3:\n",
            "      Successfully uninstalled gast-0.3.3\n",
            "  Found existing installation: tensorboard 2.2.2\n",
            "    Uninstalling tensorboard-2.2.2:\n",
            "      Successfully uninstalled tensorboard-2.2.2\n",
            "  Found existing installation: tensorflow-estimator 2.2.0\n",
            "    Uninstalling tensorflow-estimator-2.2.0:\n",
            "      Successfully uninstalled tensorflow-estimator-2.2.0\n",
            "  Found existing installation: tensorflow 2.2.0\n",
            "    Uninstalling tensorflow-2.2.0:\n",
            "      Successfully uninstalled tensorflow-2.2.0\n",
            "Successfully installed gast-0.2.2 tensorboard-2.0.2 tensorflow-2.0.0 tensorflow-estimator-2.0.1\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "display_data",
          "data": {
            "application/vnd.colab-display-data+json": {
              "pip_warning": {
                "packages": [
                  "gast",
                  "tensorboard",
                  "tensorflow",
                  "tensorflow_estimator"
                ]
              }
            }
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "stream",
          "text": [
            "Requirement already satisfied: tensorflow_hub in /usr/local/lib/python3.6/dist-packages (0.8.0)\n",
            "Requirement already satisfied: protobuf>=3.8.0 in /usr/local/lib/python3.6/dist-packages (from tensorflow_hub) (3.10.0)\n",
            "Requirement already satisfied: six>=1.12.0 in /usr/local/lib/python3.6/dist-packages (from tensorflow_hub) (1.12.0)\n",
            "Requirement already satisfied: numpy>=1.12.0 in /usr/local/lib/python3.6/dist-packages (from tensorflow_hub) (1.18.5)\n",
            "Requirement already satisfied: setuptools in /usr/local/lib/python3.6/dist-packages (from protobuf>=3.8.0->tensorflow_hub) (47.1.1)\n",
            "Requirement already satisfied: bert-for-tf2 in /usr/local/lib/python3.6/dist-packages (0.14.4)\n",
            "Requirement already satisfied: py-params>=0.9.6 in /usr/local/lib/python3.6/dist-packages (from bert-for-tf2) (0.9.7)\n",
            "Requirement already satisfied: params-flow>=0.8.0 in /usr/local/lib/python3.6/dist-packages (from bert-for-tf2) (0.8.2)\n",
            "Requirement already satisfied: numpy in /usr/local/lib/python3.6/dist-packages (from params-flow>=0.8.0->bert-for-tf2) (1.18.5)\n",
            "Requirement already satisfied: tqdm in /usr/local/lib/python3.6/dist-packages (from params-flow>=0.8.0->bert-for-tf2) (4.41.1)\n",
            "Requirement already satisfied: sentencepiece in /usr/local/lib/python3.6/dist-packages (0.1.91)\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "_V_xh0SSjDST",
        "colab_type": "code",
        "outputId": "9e411c54-93a1-4a1c-eac5-8da92c51f349",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 833
        }
      },
      "source": [
        "#imports\n",
        "import matplotlib.pyplot as plt\n",
        "import numpy as np\n",
        "import pandas as pd\n",
        "import plotly.express as px\n",
        "from nltk.corpus import stopwords\n",
        "import matplotlib as mpl\n",
        "import re\n",
        "from sklearn.feature_extraction.text import TfidfVectorizer\n",
        "from sklearn.naive_bayes import MultinomialNB\n",
        "from sklearn import model_selection\n",
        "from sklearn.model_selection import train_test_split\n",
        "from sklearn.metrics import classification_report\n",
        "import tensorflow_hub as hub\n",
        "import tensorflow as tf\n",
        "import bert\n",
        "from tensorflow.keras.models import Model\n",
        "import math\n",
        "import nltk\n",
        "\n",
        "nltk.download(\"popular\")\n",
        "mpl.style.use(\"seaborn\")\n",
        "STOP = stopwords.words('english')"
      ],
      "execution_count": 22,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "[nltk_data] Downloading collection 'popular'\n",
            "[nltk_data]    | \n",
            "[nltk_data]    | Downloading package cmudict to /root/nltk_data...\n",
            "[nltk_data]    |   Package cmudict is already up-to-date!\n",
            "[nltk_data]    | Downloading package gazetteers to /root/nltk_data...\n",
            "[nltk_data]    |   Package gazetteers is already up-to-date!\n",
            "[nltk_data]    | Downloading package genesis to /root/nltk_data...\n",
            "[nltk_data]    |   Package genesis is already up-to-date!\n",
            "[nltk_data]    | Downloading package gutenberg to /root/nltk_data...\n",
            "[nltk_data]    |   Package gutenberg is already up-to-date!\n",
            "[nltk_data]    | Downloading package inaugural to /root/nltk_data...\n",
            "[nltk_data]    |   Package inaugural is already up-to-date!\n",
            "[nltk_data]    | Downloading package movie_reviews to\n",
            "[nltk_data]    |     /root/nltk_data...\n",
            "[nltk_data]    |   Package movie_reviews is already up-to-date!\n",
            "[nltk_data]    | Downloading package names to /root/nltk_data...\n",
            "[nltk_data]    |   Package names is already up-to-date!\n",
            "[nltk_data]    | Downloading package shakespeare to /root/nltk_data...\n",
            "[nltk_data]    |   Package shakespeare is already up-to-date!\n",
            "[nltk_data]    | Downloading package stopwords to /root/nltk_data...\n",
            "[nltk_data]    |   Package stopwords is already up-to-date!\n",
            "[nltk_data]    | Downloading package treebank to /root/nltk_data...\n",
            "[nltk_data]    |   Package treebank is already up-to-date!\n",
            "[nltk_data]    | Downloading package twitter_samples to\n",
            "[nltk_data]    |     /root/nltk_data...\n",
            "[nltk_data]    |   Package twitter_samples is already up-to-date!\n",
            "[nltk_data]    | Downloading package omw to /root/nltk_data...\n",
            "[nltk_data]    |   Package omw is already up-to-date!\n",
            "[nltk_data]    | Downloading package wordnet to /root/nltk_data...\n",
            "[nltk_data]    |   Package wordnet is already up-to-date!\n",
            "[nltk_data]    | Downloading package wordnet_ic to /root/nltk_data...\n",
            "[nltk_data]    |   Package wordnet_ic is already up-to-date!\n",
            "[nltk_data]    | Downloading package words to /root/nltk_data...\n",
            "[nltk_data]    |   Package words is already up-to-date!\n",
            "[nltk_data]    | Downloading package maxent_ne_chunker to\n",
            "[nltk_data]    |     /root/nltk_data...\n",
            "[nltk_data]    |   Package maxent_ne_chunker is already up-to-date!\n",
            "[nltk_data]    | Downloading package punkt to /root/nltk_data...\n",
            "[nltk_data]    |   Package punkt is already up-to-date!\n",
            "[nltk_data]    | Downloading package snowball_data to\n",
            "[nltk_data]    |     /root/nltk_data...\n",
            "[nltk_data]    |   Package snowball_data is already up-to-date!\n",
            "[nltk_data]    | Downloading package averaged_perceptron_tagger to\n",
            "[nltk_data]    |     /root/nltk_data...\n",
            "[nltk_data]    |   Package averaged_perceptron_tagger is already up-\n",
            "[nltk_data]    |       to-date!\n",
            "[nltk_data]    | \n",
            "[nltk_data]  Done downloading collection popular\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "ckiITZZ5WWd7",
        "colab_type": "text"
      },
      "source": [
        "DATA CLEANING"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "QTTjdzpRAZNj",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "train = pd.read_csv('/content/DataScienceSS20Project/nlp-getting-started/train.csv', dtype={'id': np.int16, 'target': np.int8})\n",
        "test = pd.read_csv('/content/DataScienceSS20Project/nlp-getting-started/test.csv', dtype={'id': np.int16})"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "ivy_bCD4orQr",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "#Data Cleaning\n",
        "#all to lower case\n",
        "#remove punctiaction and other characters etc\n",
        "#remove Stop words\n",
        "\n",
        "def clean_data(x):\n",
        "  x= x.lower()\n",
        "  x = re.sub('https?://\\S+|www\\.\\S+', '', x)\n",
        "  x = re.sub(\"'\", \" \", x)\n",
        "  x = re.sub(r\"&gt;\", \"\", x)\n",
        "  x = re.sub(r\"&lt;\", \"\", x)\n",
        "  x = re.sub(r\"&amp;\", \"\", x)\n",
        "  x = re.sub(r\"\\n\", \" \", x)\n",
        "  x = re.sub(r\"\\x89ûò\", \"\", x)\n",
        "  x = re.sub(r\"\\x89ûò\", \"\", x)\n",
        "  x = re.sub(r\"\\x89ûª\", \"\", x)\n",
        "  x = re.sub(r\"åê\", \"\", x)\n",
        "  x = re.sub(r\"\\x89ûï\", \"\", x)\n",
        "  x = re.sub(r\"\\x89ûò\", \"\", x)\n",
        "  x = re.sub(r\"\\x89ûª\", \"\", x)\n",
        "  x = re.sub(r\"\\x89ûò\", \"\", x)\n",
        "  x = re.sub(r\"\\x89ûª\", \"\", x)\n",
        "  x = re.sub(r\"\\x89û\", \"\", x)\n",
        "  x = re.sub(r\"\\x9d\", \"\", x)\n",
        "  x = re.sub('\\w*\\d\\w*', '', x)\n",
        "  x = re.sub('ì©', 'é', x)\n",
        "  x = re.sub(\"[@.!?\\\\-*#+_:,;=><'§$%&/°^`´{}>|~()]\", \"\",x)\n",
        "  return x\n",
        "\n",
        "train['text']=train['text'].apply(lambda x: clean_data(x))\n",
        "test['text']=test['text'].apply(lambda x: clean_data(x))"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "x667-R4x4vGt",
        "colab_type": "code",
        "outputId": "4652fa55-588f-437c-eb7f-2e87c4385349",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 153
        }
      },
      "source": [
        "train['text'].values[:10]\n",
        "test['text'].values[:10]"
      ],
      "execution_count": 25,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "array(['just happened a terrible car crash',\n",
              "       'heard about earthquake is different cities stay safe everyone',\n",
              "       'there is a forest fire at spot pond geese are fleeing across the street i cannot save them all',\n",
              "       'apocalypse lighting spokane wildfires',\n",
              "       'typhoon soudelor kills  in china and taiwan',\n",
              "       'we re shakingit s an earthquake',\n",
              "       'they d probably still show more life than arsenal did yesterday eh eh',\n",
              "       'hey how are you', 'what a nice hat', 'fuck off'], dtype=object)"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 25
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "S3OWJyk-7Eox",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "tokenizer = nltk.tokenize.RegexpTokenizer(r'\\w+')\n",
        "\n",
        "train['text'] = train['text'].apply(lambda x: tokenizer.tokenize(x))\n",
        "test['text'] = test['text'].apply(lambda x: tokenizer.tokenize(x))"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "CI5Ug-ef62XL",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "def remove_stopwords(tokens):\n",
        "  toke=[]\n",
        "  for t in tokens: \n",
        "    if t not in STOP:\n",
        "       toke.append(t)\n",
        "  return toke\n",
        "\n",
        "train['text'] = train['text'].apply(lambda x: remove_stopwords(x))\n",
        "test['text'] = test['text'].apply(lambda x: remove_stopwords(x))"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "n61wQL1uG5Sf",
        "colab_type": "code",
        "outputId": "0329bf9f-ed76-4e11-934d-7c88f92c70f8",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 119
        }
      },
      "source": [
        "def combine_tokens(tokens):\n",
        "  a=\"\"\n",
        "  for t in tokens:\n",
        "    a = a + t + \" \"\n",
        "  return a\n",
        "\n",
        "  \n",
        "train['text'] = train['text'].apply(lambda x: combine_tokens(x))\n",
        "test['text'] = test['text'].apply(lambda x: combine_tokens(x))\n",
        "train['text'].head()"
      ],
      "execution_count": 28,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "0        deeds reason earthquake may allah forgive us \n",
              "1               forest fire near la ronge sask canada \n",
              "2    residents asked shelter place notified officer...\n",
              "3    people receive wildfires evacuation orders cal...\n",
              "4    got sent photo ruby alaska smoke wildfires pou...\n",
              "Name: text, dtype: object"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 28
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "NLJfAej_MhE2",
        "colab_type": "text"
      },
      "source": [
        "BERT Model Here"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "O3YU69YAKDqx",
        "colab_type": "code",
        "outputId": "05012aac-6486-424b-9eff-792011d18274",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 425
        }
      },
      "source": [
        "max_seq_length = 160  \n",
        "input_word_ids = tf.keras.layers.Input(shape=(max_seq_length,), dtype=tf.int32,\n",
        "                                       name=\"input_word_ids\")\n",
        "input_mask = tf.keras.layers.Input(shape=(max_seq_length,), dtype=tf.int32,\n",
        "                                   name=\"input_mask\")\n",
        "segment_ids = tf.keras.layers.Input(shape=(max_seq_length,), dtype=tf.int32,\n",
        "                                    name=\"segment_ids\")\n",
        "bert_layer = hub.KerasLayer(\"https://tfhub.dev/tensorflow/bert_en_uncased_L-12_H-768_A-12/1\",\n",
        "                            trainable=True)\n",
        "pooled_output, sequence_output = bert_layer([input_word_ids, input_mask, segment_ids])\n",
        "clf_output = sequence_output[:, 0, :]\n",
        "\n",
        "x = tf.keras.layers.Dropout(0.1)(clf_output)\n",
        "out = tf.keras.layers.Dense(1, activation='sigmoid')(x)\n",
        "#out = tf.keras.layers.Dense(1, activation='sigmoid')(clf_output)\n",
        "\n",
        "model = Model(inputs=[input_word_ids, input_mask, segment_ids], outputs=out)\n",
        "model.summary()"
      ],
      "execution_count": 29,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Model: \"model_1\"\n",
            "__________________________________________________________________________________________________\n",
            "Layer (type)                    Output Shape         Param #     Connected to                     \n",
            "==================================================================================================\n",
            "input_word_ids (InputLayer)     [(None, 160)]        0                                            \n",
            "__________________________________________________________________________________________________\n",
            "input_mask (InputLayer)         [(None, 160)]        0                                            \n",
            "__________________________________________________________________________________________________\n",
            "segment_ids (InputLayer)        [(None, 160)]        0                                            \n",
            "__________________________________________________________________________________________________\n",
            "keras_layer_1 (KerasLayer)      [(None, 768), (None, 109482241   input_word_ids[0][0]             \n",
            "                                                                 input_mask[0][0]                 \n",
            "                                                                 segment_ids[0][0]                \n",
            "__________________________________________________________________________________________________\n",
            "tf_op_layer_strided_slice_1 (Te [(None, 768)]        0           keras_layer_1[0][1]              \n",
            "__________________________________________________________________________________________________\n",
            "dropout_1 (Dropout)             (None, 768)          0           tf_op_layer_strided_slice_1[0][0]\n",
            "__________________________________________________________________________________________________\n",
            "dense_1 (Dense)                 (None, 1)            769         dropout_1[0][0]                  \n",
            "==================================================================================================\n",
            "Total params: 109,483,010\n",
            "Trainable params: 109,483,009\n",
            "Non-trainable params: 1\n",
            "__________________________________________________________________________________________________\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "GopZg71RKl7P",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "FullTokenizer = bert.bert_tokenization.FullTokenizer\n",
        "vocab_file = bert_layer.resolved_object.vocab_file.asset_path.numpy()\n",
        "do_lower_case = bert_layer.resolved_object.do_lower_case.numpy()\n",
        "tokenizer = FullTokenizer(vocab_file, do_lower_case)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "xXlBYSYJUkH2",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "# from https://www.kaggle.com/gunesevitan/nlp-with-disaster-tweets-eda-cleaning-and-bert#7.-Model\n",
        "def bert_encode(texts, tokenizer, max_len=512):\n",
        "    all_tokens = []\n",
        "    all_masks = []\n",
        "    all_segments = []\n",
        "    \n",
        "    for text in texts:\n",
        "        text = tokenizer.tokenize(text)\n",
        "            \n",
        "        text = text[:max_len-2]\n",
        "        input_sequence = [\"[CLS]\"] + text + [\"[SEP]\"]\n",
        "        pad_len = max_len - len(input_sequence)\n",
        "        \n",
        "        tokens = tokenizer.convert_tokens_to_ids(input_sequence)\n",
        "        tokens += [0] * pad_len\n",
        "        pad_masks = [1] * len(input_sequence) + [0] * pad_len\n",
        "        segment_ids = [0] * max_len\n",
        "        \n",
        "        all_tokens.append(tokens)\n",
        "        all_masks.append(pad_masks)\n",
        "        all_segments.append(segment_ids)\n",
        "    return np.array(all_tokens), np.array(all_masks), np.array(all_segments)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "VGeXL6CDUl1I",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "train_input = bert_encode(train.text.values, tokenizer, max_len=max_seq_length)\n",
        "test_input = bert_encode(test.text.values, tokenizer, max_len=max_seq_length)\n",
        "train_labels = train.target.values"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "1eb-p376RJfB",
        "colab_type": "code",
        "outputId": "eed2e6c9-a68f-43db-dabf-ea97db71d5fa",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 187
        }
      },
      "source": [
        "checkpoint =  tf.keras.callbacks.ModelCheckpoint('model.h5', monitor='val_loss', save_best_only=True)\n",
        "\n",
        "optimizer = tf.keras.optimizers.Adam(lr=0.000001)\n",
        "model.compile(loss='binary_crossentropy', optimizer=optimizer, metrics=['accuracy'])\n",
        "history=model.fit(\n",
        "    train_input, train_labels,\n",
        "    validation_split=0.2,\n",
        "    callbacks=[checkpoint],\n",
        "    epochs=5,\n",
        "    batch_size=32\n",
        ")"
      ],
      "execution_count": 33,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Epoch 1/5\n",
            "191/191 [==============================] - 125s 654ms/step - loss: 0.4679 - accuracy: 0.7824 - val_loss: 0.3939 - val_accuracy: 0.8207\n",
            "Epoch 2/5\n",
            "191/191 [==============================] - 119s 625ms/step - loss: 0.3067 - accuracy: 0.8724 - val_loss: 0.4446 - val_accuracy: 0.8089\n",
            "Epoch 3/5\n",
            "191/191 [==============================] - 119s 625ms/step - loss: 0.1753 - accuracy: 0.9348 - val_loss: 0.5105 - val_accuracy: 0.8181\n",
            "Epoch 4/5\n",
            "191/191 [==============================] - 119s 625ms/step - loss: 0.0867 - accuracy: 0.9696 - val_loss: 0.5767 - val_accuracy: 0.8070\n",
            "Epoch 5/5\n",
            "191/191 [==============================] - 119s 625ms/step - loss: 0.0606 - accuracy: 0.9765 - val_loss: 0.7089 - val_accuracy: 0.7991\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "_S-yQEAJecFn",
        "colab_type": "code",
        "outputId": "f7ec8a03-94e6-4d3e-b60a-89841da9626f",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 735
        }
      },
      "source": [
        "# summarize history for accuracy\n",
        "plt.plot(history.history['accuracy'])\n",
        "plt.plot(history.history['val_accuracy'])\n",
        "plt.title('model accuracy')\n",
        "plt.ylabel('accuracy')\n",
        "plt.xlabel('epoch')\n",
        "plt.legend(['train', 'test'], loc='upper left')\n",
        "plt.show()\n",
        "# summarize history for loss\n",
        "plt.plot(history.history['loss'])\n",
        "plt.plot(history.history['val_loss'])\n",
        "plt.title('model loss')\n",
        "plt.ylabel('loss')\n",
        "plt.xlabel('epoch')\n",
        "plt.legend(['train', 'test'], loc='upper left')\n",
        "plt.show()"
      ],
      "execution_count": 34,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 576x396 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAe8AAAFnCAYAAACPasF4AAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4yLjEsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+j8jraAAAgAElEQVR4nOzdeXyU5b028OuZmewzSWaSmezrZCUhQMIeCIuEhEXFBRutYGsVe/C8Fpe+1Via2ipVj+XYco6n1ffY9lhPjWJELEjcAIGEHQIJgewr2WeyTPZk5v0jYSRAQpBMZsn1/Xz46MwzSX63j8yV3/3ccz+CwWAwgIiIiKyGyNwFEBER0a1heBMREVkZhjcREZGVYXgTERFZGYY3ERGRlWF4ExERWRmGNxHhxRdfxI4dO8Z8TVZWFn70ox+N+3kiMh2GNxERkZVheBNZmZqaGixatAjvvPMOUlJSkJKSgrNnz2LTpk1YvHgxXnjhBeNrP//8c6xduxapqanYuHEjqqqqAABarRaPPvooli9fjk2bNqGjo8P4NSUlJXj44YeRkpKCO++8E+fPnx93ba2trfjZz36GlJQUrF69Gm+//bbx2L//+78b6924cSMaGhrGfJ6IRicxdwFEdOu0Wi2USiWys7Px1FNP4emnn8bHH38MQRCQlJSEf/mXf4FEIsHWrVvx8ccfIygoCO+++y5+9atf4a9//SveeecdyOVyvPvuu6ipqcFdd92F8PBw6PV6PPnkk3jsscewfv16nDp1Cps3b8b+/fvHVdf27dvh5uaG7OxstLa24p577kF8fDzc3Nywb98+/POf/4SdnR3ee+895ObmIiYm5obPr1u3zsT/BYmsGztvIis0MDCA1NRUAEBERASmT58OhUIBuVwOpVKJxsZGHDlyBPPmzUNQUBAAYP369Th27BgGBgZw8uRJrFq1CgDg7++PuXPnAgDKysrQ0tKC+++/HwCQkJAAhUKBM2fOjKuugwcP4qGHHgIAuLu7Izk5GUeOHIGrqys0Gg0+++wztLW1YcOGDVi3bt2ozxPR2BjeRFZILBbD0dERACASieDs7Dzi2ODgILRaLVxdXY3Py2QyGAwGaLVatLW1QSaTGY9deV17ezt6enqwatUqpKamIjU1FS0tLWhtbR1XXRqNZsTPdHV1RUtLC7y8vLBjxw7s27cPS5cuxaZNm1BXVzfq80Q0NoY3kY3y8PAYEbptbW0QiUSQy+VwdXUdcZ1bo9EAAFQqFVxcXLBv3z7jn8OHDyM5OXlcP9PT03PEz2xtbYWnpycAYP78+Xj77bdx5MgR+Pj44I033hjzeSIaHcObyEYlJibi5MmTqK6uBgB88MEHSExMhEQiwcyZM/HVV18BAKqqqnDq1CkAgJ+fH7y9vbFv3z4AQ6H+zDPPoKura1w/c+nSpcjMzDR+7ZdffomlS5fi8OHDeOmll6DX6+Hs7IyoqCgIgjDq80Q0Ni5YI7JR3t7eePnll7F582b09/fD398fv/3tbwEATzzxBJ5++mksX74carUaK1euBAAIgoDt27fj17/+Nd58802IRCL8+Mc/HjEtP5YtW7bg17/+NVJTUyESibBp0ybExcWht7cXe/bsQUpKCuzt7aFQKLBt2zaoVKobPk9EYxN4P28iIiLrwmlzIiIiK8PwJiIisjIMbyIiIivD8CYiIrIyDG8iIiIrYzUfFWtq6rj5i26BXO4MrXZ8n121dByLZbKVsdjKOACOxVLZylhMMQ6lUnbD56ds5y2RiM1dwoThWCyTrYzFVsYBcCyWylbGMpnjmLLhTUREZK0Y3kRERFaG4U1ERGRlGN5ERERWhuFNRERkZRjeREREVobhTUREZGUY3rfpwIGvx/W6P/zh97h8udbE1RAR0VTA8L4NdXWX8dVX2eN67c9+9ix8ff1MXBEREU0FVrM9qiXavv01FBYWYPHiOVi5chXq6i7jzTffwu9+9xs0NTWiu7sbjz66CYmJi/Gv/7oJzzzzf7F//9fo7NShqqoStbU1eOqpZ7FgQaK5h0JERFbEZsL7w29KcOJi47hfLxYLGBw0jPmaOVEqPLA8bNTjDz64AVlZHyIkRI2qqgq89db/g1arwdy587Fq1VrU1tZg69bnkZi4eMTXNTY24I03/oijR3Pw6acfM7yJiKxcfWcDiruLEOYYDkEQTP7zbCa8zS06OgYAIJO5orCwALt3Z0EQRGhvb7vutXFxMwEAKpUKOp1uUuskIqKJ09rbhr3lXyLn8gkAwLZFv4Sr/Y1vJjKRTBre27ZtQ15eHgRBQHp6OuLi4gAADQ0NeO6554yvq66uxrPPPos777zze/+sB5aHjdklX0uplE3oncrs7OwAAF9+uQ/t7e34z//8f2hvb8djj2247rVi8Xeb1xsMY3f/RERkeboHevBV1UF8U/Ut+vT98HJW4ccJ98PVzvTBDZgwvI8fP47KykpkZmaitLQU6enpyMzMBAB4eXnhvffeAwAMDAxgw4YNWL58ualKMRmRSITBwcERz7W2tsLHxxcikQgHD36D/v5+M1VHREQTbVA/iMOXj2Fv+ZfQ9XfC1V6G+0Puwnyf2fD2cp/w21ePxmThnZubixUrVgAA1Go12traoNPpIJVKR7zuk08+QUpKClxcXExViskEBYXg0qWL8PHxhbu7OwBg6dLleP75Z3DhQj7WrLkLKpUKf/nLO2aulIiIbofBYMDZpnzsLv0cjd3NcBDbY23ISiwPTIKD2H7S6xEMJpq33bp1K5YsWWIM8IceegivvPIKQkJCRrzugQcewLvvvntdqF9rYGDQZu75SkRE1uNiUwney8tCcUs5xIIId6gX4f6YNXB3dDVbTZO2YO1GvyOcOXMGoaGhNw1uANBquya0nom+5m1OHItlspWx2Mo4AI7FUlnqWOo7G/Fp6ec411wAAJipnI671KnwclaivwNo6hhZsynGoVTe+Bq6ycJbpVKhubnZ+LixsRFKpXLEaw4cOIAFCxaYqgQiIqJb1tbbgb3lXyCn7gT0Bj1C3YJxT9hqhLoFm7s0I5OFd2JiInbs2IG0tDQUFBRApVJd12GfP38eq1evNlUJRERE49Yz0IOvq77FV9Xfom+wD17OStytXoU4z5hJ+ez2rTBZeMfHxyMmJgZpaWkQBAEZGRnIysqCTCZDcnIyAKCpqQkeHh6mKoGIiOimBvWDyKk7jj3lX6KjTweZvRT3hq3FQp85EIssc62VSa95X/1ZbgCIiooa8fizzz4z5Y8nIiIalcFgQF5zAT4t3YvGrmbYi+2xOiQZdwQkwVHiYO7yxsQd1oiIaMopa6vAJyV7UNZWCZEgwiK/+VgdnAw3h8nZZOV2Mbxv04EDX2Pp0jvG/fqzZ08jKCgYcrnChFUREdGNNHQ1YXfp5zjblA8AmOEZg7vUq+DtojJzZbeG4X0brtwS9FbCe8+e3XjwwYcZ3kREk6i9rwN7y7/CkcvHoDfoEeIahHvC1kDtHmzu0r4XhvdtuHJL0HfffRtlZSXo6OjA4OAgtmz5OcLCwvH3v/8VBw/uh0gkQmLiYkRHT8OhQwdQXl6Gl19+Hd7e3uYeAhGRTesZ6MU31d/iq6qD6B3sg8rJE3erV2GGMtbiVpDfCpsJ76ySf+JM4/lxv14sEjCoH3tzuVmq6bg3bO2ox6/cElQkEmHevIW48851KC8vwx/+8AbefPMtfPDB37Fr1z6IxWLs2vUx5syZj7CwCDzzzP9lcBMRmdCgfhC5dSewp/xLtPd1QGrngnXq1Uj0nWexK8hvhc2EtzmdP38Ora1aZGfvBQD09vYAAJYuvQNbtmxGcnIqVq5MNWeJRERTgsFgwPnmC9hV+jkauhphL7LDquA7sCJwCRwljuYub8LYTHjfG7Z2zC75WhO5jZ2dnQRPP/1zxMbGjXj+uedeQGVlBb755kv8n//zBN5++28T8vOIiOh65W1V+KRkD0rbyiFAQKLvPKwJSYabg/n2IDcVmwlvc7hyS9Bp02Lx7bcHEBsbh/LyMhw7loO1a9fho4/+gR//+HH8+MeP4+zZM+jq6rzhbUSJiOj7a+xqwu7SfTjTNHTpdLrnNKxTr4K3i5eZKzMdhvdtuPqWoA0N9di8+THo9Xps2fIcpFIpWlu1ePzxjXByckZsbBxcXd0wc2Y8fvnLX+B3v/s9QkPV5h4CEZHV6ujT4fOKr3Co9ij0Bj2CXANwj3oNwuWh5i7N5Ex2S9CJZoo7tVjiXWy+D47FMtnKWGxlHADHYqludSy9g33YX30IX1YeQM9gLzydPHC3ehVmKaebdQW5TdxVjIiIaCIN6gdxtP4k9pR9gbbhFeTr1alY5DsPEtHUirOpNVoiIrI6BoMB+S2F2FX6Oeo7G2AnskNq0HKsCFoKJxtaQX4rGN5ERGSxKtur8UnJHhS3lkGAgIU+c7EmNBnuDm7mLs2sGN5ERGRxmrpa8FnZPpxqzAMAxHpE4271KvhKucEVwPAmIiILouvrNK4gHzQMIkgWgHVhqxEh56dzrsbwJiIis+sb7MP+6sP4ovIAegZ74OmowF3qVYhXxVn1HuSmwvAmIiKz0Rv02F+Wg3+c243W3ja42Dnj/tC7sNhv/pRbQX4r+F+GiIgmncFgwAXNJewq2YvLnfWwE0mwMmgZVgYthZPEydzlWTyGNxERTaqq9hp8UroXRdoSCBCwNGQBVvgsg9zR3dylWQ2GNxERTYrmbg0+K9uHkw1nAQAxHlG4W70KM0MibGa3uMnC8CYiIpPS9Xciu+IbHKzJwaBhEIEyP6xTr0GkIszcpVkthjcREZlE32A/DtQcxheV+9E90AMPRznuCk1FvNcMiASRucuzagxvIiKaUHqDHsfrT+OzsuyhFeQSZ9wXthaL/RfCjivIJwT/KxIR0YQwGAwo1BRhV+le1OrqIBFJkBy4FCuDlsHZjivIJxLDm4iIblt1Ry12lezFRW0xBAiY552AtaEroXCUm7s0m8TwJiKi762lW4vPyvbhRMMZAEC0IgLr1KvhL/M1c2W2jeFNRES3rLO/a3gF+REMGAYRIPXFurA1iFKEm7u0KYHhTURE49Y/2I+DtTnYV/ENuge6oXCU487QFMz2mskV5JOI4U1ERDelN+hxsuEsdpfug7a3Fc4SJ9wTtgZL/BbCTmxn7vKmHIY3ERGN6aKmGJ+U7EGN7jIkIglWBC5BStAyONs5m7u0KYvhTUREN1TTcRm7SveiUFMEAQLmesdjbUgKPJy4gtzcGN5ERDSCpkeLf5Z9geP1p2GAAVHycKwLW40AmZ+5S6NhJg3vbdu2IS8vD4IgID09HXFxccZjdXV1eOaZZ9Df349p06bhN7/5jSlLISKim+jq78YXlfuxv+YwBvQD8JP64B71GkR7RJi7NLqGycL7+PHjqKysRGZmJkpLS5Geno7MzEzj8VdffRWPPvookpOT8dJLL+Hy5cvw9eXnAomIJlu/fgDf1uQgu+IbdA50Qe7gjjtDUzDHexZXkFsok4V3bm4uVqxYAQBQq9Voa2uDTqeDVCqFXq/HqVOnsH37dgBARkaGqcogIqJR6A16nGrIw2dl+9DSo4WTxBHr1Kux1D+RK8gtnMnCu7m5GTExMcbHCoUCTU1NkEql0Gg0cHFxwe9+9zsUFBRg9uzZePbZZ01VChERXeOSpgSflO5BdUctJIIYywMWIyV4OaR2LuYujcZh0hasGQyGEf/e0NCAjRs3ws/PD5s2bcKBAwewdOnSUb9eLneGRCKe0JqUStmEfj9z4lgsk62MxVbGAXAsVa21eP/cJzhTVwAAWBQ0F2nT74LKxWOiy7sltnJeJmscJgtvlUqF5uZm4+PGxkYolUoAgFwuh6+vLwIDAwEACxYsQHFx8ZjhrdV2TWh9SqUMTU0dE/o9zYVjsUy2MhZbGQcwtcei7WnFP8u+wLH6UzDAgAh5GO5Rr0agqz/QBTR1me+/i62cF1OMY7RfBky2EiExMRHZ2dkAgIKCAqhUKkilUgCARCJBQEAAKioqjMdDQkJMVQoR0ZTVPdCNT0s/x0tHX8fR+pPwcfHC5hk/wVMzHx8KbrJKJuu84+PjERMTg7S0NAiCgIyMDGRlZUEmkyE5ORnp6el4/vnnYTAYEBERgeXLl5uqFCKiKWdAP4BDtUfxecVX6OzvgruDG9aGpmCedzxXkNsAk17zfu6550Y8joqKMv57UFAQ/vGPf5jyxxMRTTl6gx5nGs9hd+k+NPdo4Ch2xN3qVVjqvwj2XEFuM7jDGhGRjSjSluKTkj2o6qiBWBBjWcAipAbdAak9V5DbGoY3EZGVu6yrx6ele5HfchEAkKCagbvUqfB0Mu8KcjIdhjcRkZVq7W3DnrIvkFt3EgYYEO4einvC1iDINcDcpZGJMbyJiKxM90APPjj/DT67+DX69f3wcfHCOvVqxHhEQRAEc5dHk4DhTURk4QwGAxq6mlCoKcJFTRGKWsvQN9gHN3tXrA1dh/k+CVxBPsUwvImILJCuvxOXNCW4qClCoaYY2t5W4zFvZxWWqRdgrmIO7MX2ZqySzIXhTURkAQb0Ayhvq8RFTTEKNcWo6qiBAUPbSrtInJGgmoEoRQSiFeGQO7rbzK5k9P0wvImIzMBgMKCxqwmFmmIUaopQ1FqKvsE+AIBIECHMPcQY1gEyP06L0wgMbyKiSdLZ34VL2hIUthShUFM0Yircy1lpDOtw91A4ShzNWClZOoY3EZGJDOoHUd5ehULNUFhXtX83Fe4sccIsVRyiFeGIkkfAw0lu5mrJmjC8iYgmiMFgQFN3s3EqvFhbip7BXgBDU+GhbsGIVkQg2iMcgTJ/ToXT98bwJiK6DV39XbikLTV+jKulR2s8pnLyxFxFwtBUuFwNJ06F0wRheBMR3YIrU+EXNcW4qClCRXu1cSrcSeKEWcrpiFZEIEoRDg8nhZmrJVvF8CYiGsPQVHiL8fPWRdqSa6bCg4bDOgJBrpwKp8nB8CYiukZXfzeKtCXDC82K0dKjMR5TOnlgriIeUYoIRMhD4SRxMmOlNFUxvIloyhvUD6Kyo3r4I1zFqGivumoq3BEzlbHGj3HxTl1kCRjeRDQlNXe3GDvrS5oS9Az2ABiaCg9xCxwO6wgEyfwhFonNXC3RSAxvIpoSuge6cbymFMcqzqFQU4Tm7hbjMU9HBWZ7z0S0IgIR7mo423EqnCwbw5uIbNLQVHiNcaFZRXsV9AY9AMBR7IgZnjHG7lrpzKlwsi4MbyKyGc3dGuPnrS9pS9A9MDQVLkBAsGsgEgJiEegQhGDXAE6Fk1VjeBOR1eoe6EGRtnS4uy5C01VT4R6OciSoZgxNhcvVcLZz5p24yGYwvInIaugNelS21xjDunzEVLgD4jxjhvYKV0RA6eQBQRDMXDGRaTC8iciitXRrhu9xXYSL2hJ0D3QDGJoKD3INMIZ1iGsgp8JpymB4E5FF6RmeCi/UFOOitgiNXc3GY3IHd8SrpiNKEYFIeRhc7JzNWCmR+TC8icis9AY9qjpqUNhSPDwVXmmcCncQ22O6Z7RxVbjKyZNT4URgeBORGWh6tFdtkFKMrqumwgNd/Yf2CpeHI8QtEBIR36aIrsW/FURkcj0DvShuHZ4K1xShoavJeEzu4I6ZyumI9hhaFS61czFjpUTWgeFNRBNOb9CjuqPWGNZlbZUYNAwCAOzF9oj1iEb08F7hKmclp8KJbhHDm4gmhLan1RjWF7XF6OzvAjA0FR4g8zOGdYhbEKfCiW4T/wYR0ffSM9CLktay4R3NilHf1Wg85u7ghgU+cxCtCEekPBxSe06FE00khjcRjYveoEdNx+XhhWbXTIWL7BDjEWXsrr2cVZwKJzIhhjcRjUrb04r8svM4XnkOl7Ql0PV3Go+NnAoPhh2nwokmDf+2EZFR72AfirWluKgtRqGmGPWdDcZjbvaumO89e2gqXBEOmb3UjJUSTW0mDe9t27YhLy8PgiAgPT0dcXFxxmPLly+Ht7c3xOKh7QzfeOMNeHl5mbIcIrqG3qBHje4yLg5vkFLWVoGB4alwO5EdpnlEYnbAdATYB8LHxYtT4UQWwmThffz4cVRWViIzMxOlpaVIT09HZmbmiNe88847cHHhQhaiyaTtaTXuFX6jqfAoeTiiFREIdR+aCueduIgsj8nCOzc3FytWrAAAqNVqtLW1QafTQSrlVBvRZDJOhQ8H9tWrwjkVTmSdTBbezc3NiImJMT5WKBRoamoaEd4ZGRmora1FQkICnn32WU7JEU2Aq1eFX9QUo7Stwrgq/MpUePTwXuHeXBVOZJUmbcGawWAY8fipp57C4sWL4ebmhieffBLZ2dlITU0d9evlcmdIJBN7uz+lUjah38+cOBbLNFljaenS4lx9IfIaCnG+4SI6enXGYyHyAMzwnoY4ryhEeqphJ7a75e/Pc2KZOBbLM1njMFl4q1QqNDd/dyu/xsZGKJVK4+N169YZ/z0pKQlFRUVjhrdW2zWh9dnSdTyOxTKZcixjTYW7O7hhvs9sRMuvnwpv1fQA6Lmln8VzYpk4FstjinGM9suAycI7MTERO3bsQFpaGgoKCqBSqYxT5h0dHdiyZQv+67/+C/b29jhx4gRSUlJMVQqR1RvPBilRinBOhRNNESYL7/j4eMTExCAtLQ2CICAjIwNZWVmQyWRITk5GUlISfvCDH8DBwQHTpk0bs+smmopG2ysc4AYpRFOdSf/GP/fccyMeR0VFGf/9kUcewSOPPGLKH09kVcY1Fa6IQKQ8jKvCiaY4/rpOZCbj3Ss8ShHOqXAiGoHhTTSJxpoKD5T5IYpT4UQ0Dnx3IDKhm902k1PhRPR9MLyJJpDeoEd1R+3Q5ij5ZbjYVMrbZhLRhGN4E92m0abCBQgIkPkOT4VHIMQtiFPhRDQh+E5CdIuungov1BSj4Zqp8AU+cxClCEdi2Ez0Wv++E0RkgRjeRDdxZSr8Snc92qrwa6fCXR1laOpgehPRxJuS4d3XP4iDp2sQrHSGs+Ot7/NMto9T4URkyabku86l6lb8+4d5kMsc8EhqFOLUHuYuicxsPFPh0YpwRMrDIbXnPeiJyLymZHjHBCvwUEoUMr+8hDc/ysOi6T5IuyOMXfgUcrOp8FiPKGN37eWs5KpwIrIoUzK8RSIBD66MRISvDO/uKcTh83UoqNCwC7dxQ1PhQ7uZXdKWXDMV7mfczYxT4URk6ab0O1Sglwy/fGQ29h6txGdHKtiF25iegV4Ut17ZK5xT4URkO6Z0eAOARCzCXYkhmBWuxH/vuXBVFx6JOLWnucujWzDmVLjYnlPhRGQzpnx4XxGgkuKXG6/uws8hcbo3HrwjnF24BdP0aI134bqkKUHnwPVT4dHDU+ESToUTkY3gu9lVru3Cj5yvx4UKLbtwC3JlKnyou75+KnyhcmiDFE6FE5EtY3jfALtwy8GpcCKi6zG8R3GjLrygXIMfrYpiF25inAonIhob3/lugl24aekNetR3NqKyowaV7dUoPVGGyx0NxuNyB3fMUMYgShGBSEUYpHacCiciYniPA7vwiWEwGKDpaUVlRzUq24f+VHXUoHewz/gaB4kDYj2ijZ+55lQ4EdH1GN634EoX/vnRSuxmF35Tur7OEUFd2V6Djn6d8bgAAV4uKgTJ/BHsGoAg1wDMDI6AVtNtxqqJiCwfw/sWScQi3JkYgpnhSry7p9DYhT+SGoUZYVO3C+8d7EN1R60xqCvaq9HSoxnxGrmDO2YppyNoOKgDZH5wkjiOeI1EzP8liYhuhu+U31OASooXNyYYu/A/7DyHxFhvpK0Ih4uNd+GD+kFc7qxHRXs1qoaDuq6zAQYYjK9xkThjmiJyOKj9EeQaAFd7mRmrJiKyHQzv23BdF55fb9wj3Va6cIPBgKbu5uGgrkFFezVqdLXo1w8YX2MvskOoW5Cxow52DYCHo4LXqomITIThPQGMXfixKuw+XG7VXXhbb/uIjrqyowbdA99dgxYJIvi6eBtDOsg1AN7OKohFYjNWTUQ0tTC8J4hELMKdC4MxK8wT/20lXXj3QDcq24c+onXlo1qtvW0jXqNy8kSMRySCXQMR5OoPf6kv7MX2ZqqYiIgAhveE87fQLrx/sB81urrhoB5aVNbQ1TTiNa72Mkz3nDbUUcsCEOjqDxc7ZzNVTEREo2F4m4C5u3Djxift1ajoGJoCr9XVG7cVBQBHsSMi5GHGqe8gmT/cHdx4nZqIyAowvE3oRl34wlhvPDiBXbjBYEBjZwvONF5ERXsVqtprrtv4RCKIESDzG1r1LRsKa5WzJ0SCaEJqICKiycXwNjFjFx4+1IXn5Nfjwm104Vc2Pqlo/27zE11/p/G4AAHeLqrhbnpoUZmv1Jt7gBMR2RC+o08Sf6UUL264tS68Z6B3aOOTju92KLt24xOFoxzzvSPhbe+N4OGNTxyv2fiEiIhsC8N7Et2oC79yLXx6qBy1nXXfrf6+0cYnds6Y5hFp7KiDXAMgs5dCqZShqanDjCMjIqLJxPA2A19PZzx+XyA+O5uHMzWF+POFgxBX6mAQvltQNrTxSfBwSPsjyDUQHo5yLigjIiKG92Ro7W0zTntXXrPxidgLgEHAYJcMkl45EsOisTgsmhufEBHRqEwa3tu2bUNeXh4EQUB6ejri4uKue83vf/97nD17Fu+9954pS5k0Xf3dqOr4buq7sqPm+o1PnD0R6xFl3E7Ux9kbX5+ow6eHy/FFiQEdMVo8lOwJF0eGNxERXc9k4X38+HFUVlYiMzMTpaWlSE9PR2Zm5ojXlJSU4MSJE7Czs64tRK8Y2vjk8vDK7xpUdlShsat5xGvc7GWI84z57gYdMn8432Djk7ULgzFz+Fp4bkE9LlRq8EhKFGaGW+bubEREZD4mC+/c3FysWLECAKBWq9HW1gadTgepVGp8zauvvoqnn34a//Ef/2GqMibMlY1PKq7aoaxWVwe9QW98jZPEEZHysBH7frs7uI37Z/grpfjlxgTsO1aFTw+X448fn8OCGG88lGx9e6QTEZHp3HJ49/X1oaWlBT4+PmO+rrm5GTExMcbHCoUCTU1NxvDOysrC3Llz4efnd6slmJzBYICmRzsiqKs6atF39cYnIgmCZP4IvCajS4IAACAASURBVBLUMn8oJ2DjE7FIhDULgjEjzBPvsgsnIqIbGFd4//nPf4azszPuv/9+3HfffXBxcUFiYiK2bNky7h9kMHz3kafW1lZkZWXhL3/5CxoaGsb19XK5MySSib0GrFQO3V+6vacDJZpKlGgqUDr8z45enfF1AgT4u/lArQhCmCIYYYpgBLr5QiI23ZIBpVKGNyO9kHWgBP+bfQl//PgcliX4Y9O66ZA6X39jkCtjsQUci+WxlXEAHIulspWxTNY4xpU++/fvxz/+8Q/s2rULy5Ytw89//nNs3LhxzK9RqVRobv7u+m9jYyOUSiUA4OjRo9BoNPjhD3+Ivr4+VFVVYdu2bUhPTx/1+2m1XeMpdVz6B/txtv0sztVeQmV7NVp6tCOOezjKEa+KM+5SNrTxicN3LxgEtJpuTIalcT4I85Hh3T2F2H+qBqcvNV7XhdvS57w5FstjK+MAOBZLZStjMcU4RvtlYFzhLZFIIAgCvv32W2No6/X6Mb8mMTERO3bsQFpaGgoKCqBSqYxT5qmpqUhNTQUA1NTU4IUXXhgzuCdaUWsp/pr3EQBAaueCGI8oBMn8jau/ZfbSm3yHyeWvHNojfeS1cC88uCICUideCycimmrGFd4ymQybNm1CfX09Zs2ahf379990s5D4+HjExMQgLS0NgiAgIyMDWVlZkMlkSE5OnpDiv69oRQQylj0NcY8DFFay8cmVa+Ezw66sSG/AhQotHkmNQrKNTDcREdH4CIarL0aPoqurCzk5OYiPj4dCoUBOTg6Cg4Ph6+s7GTUCgEmmIqx1mmZQrzd24QODBixN8Me9i0Jsogu35vNyLVsZi62MA+BYLJWtjGUyp83HtTRao9FALpdDoVDgww8/xD//+U90d0/ONV+63pUuPONHcxDiI8OBUzXY+v+O4Uxxk7lLIyKiSTCu8H7hhRdgZ2eHCxcu4KOPPkJKSgpefvllU9dGN+GnlCJ9QwI2ro5GZ08/dnx8Hu98VgBdd7+5SyMiIhMaV3gLgoC4uDh8+eWX+OEPf4glS5ZgHLPtNAnEIhHW3xFh7MJzCxrYhRMR2bhxhXdXVxfOnTuH7OxsJCUloa+vD+3t7aaujW7BlS78viWh7MKJiGzcuML70UcfxdatW/GDH/wACoUCO3bswNq1a01dG90i47XwH89lF05EZMPGtdr8itbWVgiCAFdX10n/eBVXm4/uRmMZ1OuRfbwauw6VYWDQgPkxXnjICj4XbuvnxRrZyjgAjsVS2cpYLG6TllOnTuEXv/gFOjs7odfrIZfL8W//9m+YPn36hBZJE0csEmH1/KDhPdIv4OiVz4WnRGJWhNLc5RER0W0Y17T59u3b8dZbbyE3NxfHjh3D9u3b8eqrr5q6NpoAfp4uSN+QgPuXqtHV048dWefxNq+FExFZtXF13iKRCBEREcbH06ZNg1g8sTcJIdMZ2YUXsgsnIrJy4+q8RSIRsrOzodPpoNPpsHfvXoa3FRrqwuPZhRMRWblxdd4vvfQSfvvb32Lr1q0QBAEzZszAb37zG1PXRibALpyIyPqNGd4PPfSQcVW5wWBAWFgYAECn0+H555/H+++/b/oKySSudOFXVqTvyDqP+dO88FCy5a9IJyKa6sYM7y1btkxWHWQG13XhFxpwoZJdOBGRpRszvOfOnTtZdZAZXenCvzhejU8OlbMLJyKycOO65k22TywSYdVwF/7fV3XhG1MiEc8unIjIooxrtTlNHb7DXfj6pWp09QzgP7LO4+3dXJFORGRJ2HnTda7uwt/dyy6ciMjSsPOmUfl6uiD94QSsX8YunIjIkrDzpjGJRAJWzQvCDDW7cCIiS8HOm8blRl34n9mFExGZBTtvGrdru/BjFxpQWKHBhpQoJESyCycimizsvOmWjejCewfxn5+wCycimkzsvOl7udKFzxzenY1dOBHR5GHnTbfFx8MFL1zThf/p03x0dPWZuzQiIpvFzptu27Vd+PHCRlys1LILJyIyEXbeNGGudOEPLAtjF05EZELsvGlCiUQCUucFYkaYxzVdeCQSIlXmLo+IyCaw8yaTuL4Lz2cXTkQ0Qdh5k8mM6ML3sgsnIpoo7LzJ5Hw8XPDCD4e68O4+duFERLeLnTdNCnbhREQTh503TSp24UREt4+dN026G3XhhZVabFgZiVVKmbnLIyKyeCYN723btiEvLw+CICA9PR1xcXHGYx9++CF27twJkUiEqKgoZGRkQBAEU5ZDFuZKF/7FiWp8cqgMb+3Kx5nSFty5IAjeCmdzl0dEZLFMFt7Hjx9HZWUlMjMzUVpaivT0dGRmZgIAuru7sWfPHrz//vuws7PDxo0bcebMGcTHx5uqHLJQV3fhf/n8InLP1+F4QT2SZvjirkUhcHOxN3eJREQWx2ThnZubixUrVgAA1Go12traoNPpIJVK4eTkhL/97W8AhoJcp9NBqeQ2mlPZUBcej+J6Hf6yOx/7z9QiJ78eKXMDkDovEI72vMJDRHSFyRasNTc3Qy6XGx8rFAo0NTWNeM3bb7+N5ORkpKamIiAgwFSlkJUQBAGJcb747WPzsGFlBBzsRNh9pALP/ykX35yuwcCg3twlEhFZhElrZwwGw3XPbdq0CRs3bsTjjz+OhIQEJCQkjPr1crkzJBLxhNaktKHFUbY0Fh9vNzzg7Ya1S8Lw6cFSZB0owd+/KMI3p2uxcfU0LIzzsZr1EbZyXmxlHADHYqlsZSyTNQ6ThbdKpUJzc7PxcWNjo3FqvLW1FcXFxZgzZw4cHR2RlJSE06dPjxneWm3XhNanVMrQ1NQxod/TXGx5LCvi/TAnUondR8px8MxlvPo/JxDq64r1S9WIDJSP8Z3Mz1bOi62MA+BYLJWtjMUU4xjtlwGTTZsnJiYiOzsbAFBQUACVSgWpVAoAGBgYwPPPP4/Ozk4AwPnz5xESEmKqUsjKubnYY8PKSLz8+DzMjlSi7HI7XvvfM/jDR3mobdKZuzwioklnss47Pj4eMTExSEtLgyAIyMjIQFZWFmQyGZKTk/Hkk09i48aNkEgkiIyMxB133GGqUshGeCucsfme6Si93IaP9pcir7QF58pakDjdB+sWhUDh6mjuEomIJoVguNHFaAtkiqkIW5imAabmWAwGA/JKW/DxgVLUNnfCTiJC8uwArJ4fBGdHy1iZbivnxVbGAXAslspWxjKZ0+aW8S5HdIsEQcDMME/EhXrgyPk67Dpcjr1HK3HwbC3uXBiMZfH+sJNw918isk18dyOrJhIJWDzDF9s2zcd9S0KhNxjwwTclePGdo8gtqIfeOiaWiIhuCcObbIKDnRhrFgTjtZ8uxMo5AWjV9eKdzy7gN389gYJyjbnLIyKaUAxvsilSJzuk3RGOVx6fj/kxXqhq0OH3mWfx+w/OoLLe+q+pEREBvOZNNkrp7oRNd8YgZU4gdh4oQUGFFgV/PYH5MV64d3EoPN2dzF0iEdH3xvAmmxbkLcOzabNQUK7BR/tLcLSgAScvNmJ5vD/WLgyG1MnO3CUSEd0yhjdNCTEhCkQHz8GxCw3IOliGL05U49C5OqyeH4jk2QGwt5vYrXeJiEyJ4U1ThkgQsCDGG7MjVdh/ugaf5VTg44Nl+OZ0LdYtCkHidB+IRNaxZzoRTW1csEZTjp1EhJVzA/HaTxdg9fwg6Lr78ZfPLyLj3eM4W9J8w5voEBFZEnbeNGU5O9rh/qVqLI/3w67D5Thyvg5/3HkOEQHuWL9MDbWvm7lLJCK6IXbeNOUpXB3x6OpovPToXMxQe6CouhWv/M8pvPXJeTRoJvZudkREE4GdN9Ewf6UUP1s/A5eqtPhwfylOXmrCmeJmJM30xV2JIXBzsTd3iUREANh5E10nMlCOX25MwOZ1sfBwc8T+07V4/s+5+PRwOXr6BsxdHhERO2+iGxEEAbOjVJgZ7olv8y5j9+FyfHq4HPvP1OLuxGAsnuELiZi/+xKRefDdh2gMErEIy+P98bsnFuCuxGD09g3ivS+KsPW/j+PkxUauTCcis2DnTTQOTg4SrFscimWz/LD7SAUOnr2Mt3blQ+3rivXLwhAR4G7uEoloCmHnTXQL3KQO2JASiZcfn4eESCVKL7fj1fdP4487z6G2udPc5RHRFMHOm+h78FY448l7pqO0tg0f7S/B2ZJm5JU2Y9F0H6xbHAq5zMHcJRKRDWN4E90GtZ8bfvHDeOSVtGDnwVIcOleHYxcakDwnABvWxJi7PCKyUQxvotskCAJmhntiulqBI+frsetQGfbkVuLbvDqsWRCEZbP8YCfhFSoimjh8RyGaIGKRCEkzfPG7JxbgviWhGNTr8cHXxXjxnaM4WlAPPVemE9EEYXgTTTAHOzHWLAjG2y+sQPLsAGg7evH2Zxfw27+eREGFxtzlEZENYHgTmYib1AEPrgjHtk3zMX+aFyobOvD7D87i95lnUdXQYe7yiMiK8Zo3kYkp3Z2w6a4YpMwNxEcHSlBQrsGFcg3mx3jhnqRQeLo5mbtEIrIyDG+iSRLkLcNzabOQX96Cj/aXIregAScuNmJ5vD/WLgyG1MnO3CUSkZVgeBNNstgQD0wLVuBYQQOyvi3DFyeqcejc0Mr0FQn+sLcTm7tEIrJwDG8iMxAJAhbEemN2lBLfnK7FP3MqsPNAKb4+VYN1i0OQGOsDkUgwd5lEZKG4YI3IjOwkYqTMDcRrP12AVfMDoevux1/2XkTGX44jr6SZNz4hohti501kAZwd7bB+aRjuiPfHrkPlOHK+Dn/YeQ6RAe5YvywMob6u5i6RiCwIO28iC6JwdcSja6Lx0k/mIk7tgUvVrXj5f07irV35aNB2mbs8IrIQ7LyJLJC/Uoot62fgUpUWH+4vwcmLjThT1IQlM31xV2IIXF3szV0iEZkRO28iCxYZKMcvN87Gv6yLhYebI745XYtf/DkXuw+Xo6dvwNzlEZGZsPMmsnCCIGBOlAqzwj1x8Oxl7D5Sjl2Hy/HNmVrcvSgEi+N8IBHz93CiqcSk4b1t2zbk5eVBEASkp6cjLi7OeOzo0aPYvn07RCIRQkJC8Morr0Ak4hsQ0WgkYhHuSPDHwlhvZB+vQvbxaryXfQlfnKjG/UtCER+hhCDw42VEU4HJ0vL48eOorKxEZmYmXnnlFbzyyisjjv/qV7/CH//4R3zwwQfo7OzEoUOHTFUKkU1xcpBg3eJQvPrEfCyd5YcmbTf+85N8bPv7KRRVt5q7PCKaBCYL79zcXKxYsQIAoFar0dbWBp1OZzyelZUFb29vAIBCoYBWqzVVKUQ2yU3qgI0pkfjtY3OREKFEaW07Xn3/NP648xwuN3eauzwiMiGThXdzczPkcrnxsUKhQFNTk/GxVCoFADQ2NuLIkSNYsmSJqUohsmk+Hi548t7pSN+QgHB/N5wtacbW/z6Gv35eCG1Hr7nLIyITmLQFazfaKaqlpQU//elPkZGRMSLob0Qud4ZEMrF7PiuVsgn9fubEsVimyRyLUinD/Bl+OF5Qj7/tvYBv8+pw9EIj7k4KxX3LwuFyGzc+4TmxTByL5ZmscZgsvFUqFZqbm42PGxsboVQqjY91Oh0ef/xxbNmyBYsWLbrp99NO8AYVSqUMTU22cU9ljsUymWssoV5S/OqR2Thyvh67DpXho6+L8XlOBe5cGIyls/xgJ7m1CTeeE8vEsVgeU4xjtF8GTDZtnpiYiOzsbABAQUEBVCqVcaocAF599VU88sgjSEpKMlUJRFOWWCRC0gxf/O6JBbg3KRSDej3+8XUxXnznKI5eqIeee6YTWTWTdd7x8fGIiYlBWloaBEFARkYGsrKyIJPJsGjRIuzatQuVlZXYuXMnAGDt2rX4wQ9+YKpyiKYkBzsx1i4MxpKZvvgspwL7T9fi7d0XkH28GuuXqjEtWGHuEonoexAMVnLbIlNMRdjCNA3AsVgqSxxLY2s3Pvm2DMcuNAAAYkMUuH+pGoFeo1+ns8RxfF8ci2WylbFM5rQ5d1gjmkJU7k544q4YpMwNwEf7S5FfrkFBuQbzY7xxT1IIPN2czF0iEY0Dw5toCgr2dsVzaTNRUK7BRwdKkVtQjxMXG3BHgj/WLAiG9DZWphOR6TG8iaYoQRAQG+qBaSEKHC2oxyffliH7eDUO5dVhzYIg3JHgD3u7if14JhFNDG4mTjTFiQQBC2N9sG3TfDywLAyCAHx0oBQvvH0Uh8/VYVBvFctiiKYUdt5EBACwk4iROi8Qi2f4YG9uJb48WYN39xbiq9M1WDzdB3OjVZA58z7iRJaA4U1EI7g42mH9sjDckeCPTw6VITe/Hu/Xd+CDr4sRp/bAwlhvxKk9b3mzFyKaOAxvIrohhasjfrJmGjbdOwN7D5UiJ78eZ4qbcaa4GS6OEsyJ9sLCWG+ofV15K1KiScbwJqIxKVwdkTI3EClzA1HdqENufj1yC+px4EwtDpyphUruhIUx3lgQ6w2lOz9qRjQZGN5ENG4BKikClofhvqWhKKzQIqegHqcvNWHX4XLsOlyOCH83LIj1xpwoFZwd+XEzIlNheBPRLROLRIgN9UBsqAe6Vw7g1KUm5OTX4VJVK4pq2vD+l8WYFe6JhbHeiAlRQCLm9XGiicTwJqLb4uQgwaI4HyyK80FLWw+OXqhHTn49TlxsxImLjZA522HeNC8kxvog0EvK6+NEE4DhTUQTxsPNEWsWBGP1/CBU1HcgJ78exy404KuTNfjqZA38PF2wINYb86d5QeHqaO5yiawWw5uIJpwgCAjxcUWIjyt+sDwM58takJtfj7Mlzdh5oBQfHyhFdLAcC2K8kRCphKM934qIbgX/xhCRSUnEIswKV2JWuBKdPf04UdiInIJ6XKjQ4kKFFu99cQkJESosjPVGdJAcIhGn1YluhuFNRJPGxdEOS2f5YeksPzRou5CbP3R9PLdg6I+71B4Lhj925q+UmrtcIovF8CYis/CSO2Pd4lDcvSgEJbVtyMmvx/HCRnx+rAqfH6tCoJcUC2N9MG+aF9xcuC0r0dUY3kRkVoIgINzfHeH+7nhoRTjySlqQk1+P82Ut+ODrYnz4TQliQxVYGOuNmWGevNMZERjeRGRB7CRizI5SYXaUCu1dfTh+oQE5+fU4V9qCc6UtcHIQY3bk0PXx8AB3iPixM5qiGN5EZJFcne2xYnYAVswOwOXmTuQWDF0fP3SuDofO1cHTzRELYryxMNYbXgpnc5dLNKkY3kRk8Xw9XXDfEjXuSQrFpUotcvLrcbKoCZ/lVOCznAqofV2xMNYbc6K9IHXitqxk+xjeRGQ1RIKA6GAFooMVeLhvEKeLm5CTX48LFRqUXm7H/35VjBlhnsO3LfXgtqxksxjeRGSVHOzFQx8ri/GGtqMXxy404Eh+HU4XNeF0URNcHCWYO23otqWhPrxtKdkWhjcRWT25zAGp8wKRMjcA1Y065OTX4+iFBuw/XYv9p2vhpXDGwhgvrEkKA3txsgUMbyKyGYIgINBLhkAvGdYvU6OgXIvcgnqcLmrCJ4fK8cmhckQGuGNhrDdmR6ng5MC3QLJO/D+XiGySWCRCnNoDcWoPdPcO4OTFRpwoakJ+aQsuVbfi718WDd+21AcxIXKIRezJyXowvInI5jk5SLB4hi/uXRGJwuJG5A5/fvx4YSOOFzbC1cUe84evjweoeNtSsnwMbyKaUjzdnXDnwmCsXRCE8roO5OTX4diFBnxxohpfnKiGv/LKbUu9IZc5mLtcohtieBPRlCQIAkJ9XRHq64q0O8JxvnRoW9azJc34aH8pdh4oxbRgBRbGeCM+QgkHe27LSpaD4U1EU55ELMKsCCVmRSih6+7HicIG5BTUo6Bcg4JyDRzsxJgdqcSCWG9EBfK2pWR+DG8ioqtIneywLN4fy+L90aDpMt6y9Ej+0B+5zMF421I/Txdzl0tTFMObiGgUXgpn3JMUirsXh6Ckpg05+XU4cbERe49WYu/RSgR5y7Aw1hvzor3gytuW0iRieBMR3YRIEBAR4I6IAHc8tCICZ0uakZNfj/wyDf5RP3zb0hAFFk73wcwwD9hJeH2cTIvhTUR0C+ztxJgb7YW50V5o6/zutqV5pS3IK22Bk4MEc6NVWBDjjXB/N37sjEzCpOG9bds25OXlQRAEpKenIy4uznist7cXv/rVr1BcXIysrCxTlkFEZBJuLvZInhOA5DkBqG3SIaegHkcLGnDw7GUcPHsZSvfvbluqkvO2pTRxTBbex48fR2VlJTIzM1FaWor09HRkZmYaj7/++uuIjo5GcXGxqUogIpo0fkop1i8Nw31JahRWaZGbX49Tl5qw+0gFdh+pQJif2/BtS1VwceRtS+n2mCy8c3NzsWLFCgCAWq1GW1sbdDodpFIpAODpp59Ga2srdu/ebaoSiIgmnUgkICZYgZhgBR5eOYDTRUO3LS2s0KKktg3/+1WR8bal00N521L6fkwW3s3NzYiJiTE+VigUaGpqMoa3VCpFa2urqX48EZHZOdpLsDDWBwtjfaBp78Gx4evjpy414dSlJkid7DAv2gsLp3sj2FvG6+M0bpO2YM1gMNzW18vlzpBM8ApOpVI2od/PnDgWy2QrY7GVcQDmG4tSKUOkWokNa2NQVtuGb05V49vTtfj6dA2+Pl0Df5UUy2cHYEm8/7ivj/O8WJ7JGofJwlulUqG5udn4uLGxEUql8nt/P622ayLKMlIqZWhq6pjQ72kuHItlspWx2Mo4AMsZi6uDGOsWBmPtvEAUlGuGb1vajP/ZW4j39hYiMtAdC2K9MTty9NuWWspYJoKtjMUU4xjtlwGThXdiYiJ27NiBtLQ0FBQUQKVSGafMiYhoaFvWGWGemBHmia6efpy8NHR9/GJVKy5WteL9L4oQH6HEwlhvRAfztqX0HZOFd3x8PGJiYpCWlgZBEJCRkYGsrCzIZDIkJyfjqaeeQn19PcrLy7FhwwY88MADuPPOO01VDhGRRXN2tEPSDF8kzfBFU2s3cgvqkZtfj6MXGnD0QgPcXOwxP8YLC2N9EKBiIzTVCYbbvRg9SUwxFWEL0zQAx2KpbGUstjIOwPrGYjAYUHa5ffje4w3o7BkAAPgrpYiPUsFBIkAuc4BC5gh3mQPkUgfYSayvO7e28zIam5g2JyKi2yMIAtR+blD7uSHtjnCcK21BTn4dzpW2YPehsht+jczZDnKpA+QyB8hdHSGX2kMuc4Tc1cH4/GjX0cl68AwSEVkBO4kICZFKJEQq0dUzgH5BQFmVBq0dvdB09EJ71Z8GbTeqGnWjfi8nBzHcpQ5QyBwgH+7aFTIH4z/lMgdInez40TULxvAmIrIyzo4SKJUyuDnc+OOzBoMB3b2D0Hb0jAh1TUcvWnW90LT3QtvRg7qW0T/FIxGLIJcNd+3DgS4fnpqXuw5N1bu52PPe5mbC8CYisjGCIMDZUQJnRyn8lKMvbuvrH4RW1wtte+/QPzuu/vceaDp6UVzditEWRgkC4C69KtRlDiOm569M2/MuaxOP4U1ENEXZ24nhJXeG1xibwgwM6tHe2TfUtRun6Ed29JX1HSjTt4/6PaROdtdNy7sPL7STyxzgInM0xfBsGsObiIhGJRGLoHB1hMJ19IDVGwzQdfVfFeg9xo7+ylT9za7DO9qLR07PXz1dPzxVL+N1eCOGNxER3RaRIMDVxR6uLvYI8r7xR5uM1+GHp+Svnqrv7B1EQ0sntB29N70O7y61H+reXR2/m56/6o+b1H5KbGbD8CYiIpP77jq8BH6eLiOOXf356L7+QbTqrlpgd91q+h4U17bBUNM2ys8Zus+6XOZ43VT91X+s/To8w5uIiCyGvZ0YKrnzmDdnGdTr0abrG3HdfSjse4xhX93YgfK6sa/DXxvoV0/ZKyz88/CWWxkREdENiEU3vw5vMBjQ0d0/ND3f8d0K+qvDvrG1G9VjXId3sBcPde9XPhNvXEn/3fV4qbMdRGa4Ds/wJiIimyMIAlyd7eHqPPp1eADo7h24anp+KNyvnaof+zq8YPy4XGSQAncnBk3KNXeGNxERTVlODhL4OVx/Hf5qV1+Hv36qfmg1fUltG2qadFg52x9SJ4Y3ERGRWY33OryHhxSt2tG79Ilk++vpiYiITEwsEk3qCnaGNxERkZVheBMREVkZhjcREZGVYXgTERFZGYY3ERGRlWF4ExERWRmGNxERkZVheBMREVkZhjcREZGVYXgTERFZGYY3ERGRlREMBoPB3EUQERHR+LHzJiIisjIMbyIiIivD8CYiIrIyDG8iIiIrw/AmIiKyMgxvIiIiKyMxdwGTYdu2bcjLy4MgCEhPT0dcXJzxWE5ODrZv3w6xWIykpCQ8+eSTZqz05sYay/Lly+Ht7Q2xWAwAeOONN+Dl5WWuUm+qqKgImzdvxo9+9CM8/PDDI45Z03kZaxzWdk5ef/11nDp1CgMDA3jiiSewcuVK4zFrOifA2GOxlvPS3d2N559/Hi0tLejt7cXmzZuxbNky43FrOic3G4u1nJOr9fT0YO3atdi8eTPuvfde4/OTcl4MNu7YsWOGTZs2GQwGg6GkpMTwwAMPjDi+atUqw+XLlw2Dg4OGBx980FBcXGyOMsflZmNZtmyZQafTmaO0W9bZ2Wl4+OGHDb/85S8N77333nXHreW83Gwc1nROcnNzDY899pjBYDAYNBqNYcmSJSOOW8s5MRhuPhZrOS979uwxvP322waDwWCoqakxrFy5csRxazonNxuLtZyTq23fvt1w7733Gj7++OMRz0/GebH5afPc3FysWLECAKBWq9HW1gadTgcAqK6uhpubG3x8fCASibBkyRLk5uaas9wxjTUWa2Nvb4933nkHKpXqumPWdF7GGoe1mTNnDv7whz8AAFxdXdHd3Y3BwUEA1nVOgLHHYk1Wr16Nxx9/HABQV1c3ohO1tnMy1lisUWlpKUpKSrB06dIRz0/WebH5afPm5mbExMQYHysUCjQ1NUEqlaKpxiY7GwAABYJJREFUqQkKhWLEserqanOUOS5jjeWKjIwM1NbWIiEhAc8++ywEQTBHqTclkUggkdz4fz9rOi9jjeMKazknYrEYzs7OAICdO3ciKSnJOIVpTecEGHssV1jLeQGAtLQ01NfX409/+pPxOWs7J1fcaCxXWNM5ee2117B161bs2rVrxPOTdV5sPryvZbCh3WCvHctTTz2FxYsXw83NDU8++SSys7ORmppqpuoIsM5z8tVXX2Hnzp149913zV3KbRttLNZ2Xj744AMUFhbi5z//OXbv3m3RoXYzo43Fms7Jrl27MHPmTAQEBJitBpufNlepVGhubjY+bmxshFKpvOGxhoYGi57+HGssALBu3Tp4eHhAIpEgKSkJRUVF5ijztlnbeRmLtZ2TQ4cO4U9/+hPeeecdyGQy4/PWeE5GGwtgPeclPz8fdXV1AIDo6GgMDg5Co9EAsL5zMtZYAOs5JwBw4MABfP3113jggQfw0Ucf4a233kJOTg6AyTsvNh/eiYmJyM7OBgAUFBRApVIZp5n9/f2h0+lQU1ODgYEB7N+/H4mJieYsd0xjjaWjowM/+clP0NfXBwA4ceIEwsPDzVbr7bC28zIaazsnHR0deP311/HnP/8Z7u7uI45Z2zkZayzWdF5OnjxpnDVobm5GV1cX5HI5AOs7J2ONxZrOCQC8+eab+Pjjj/Hhhx9i/fr12Lx5MxYuXAhg8s7LlLir2BtvvIGTJ09CEARkZGTgwoULkMlkSE5OxokTJ/DGG28AAFauXImf/OQnZq52bGON5W9/+xt27doFBwcHTJs2DVu3brXY6bX8/Hy89tprqK2thUQigZeXF5YvXw5/f3+rOi83G4c1nZPMzEzs2LEDISEhxufmzZuHyMhIqzonwM3HYi3npaenBy+++CLq6urQ09ODf/3Xf0Vra6tVvn/dbCzWck6utWPHDvj5+QHApJ6XKRHeREREtsTmp82J/n9798/SSBSFYfwRNCZCMCI4kkostDEIAWOXwo+QUlFSBKyF2InNkKAEVFDbgMQR00wraKMpFBshAUVQIcQgiCj4v5JssSDLLlssy264zvvr5lb3VC/nDJwrIvLVKLxFREQMo/AWERExjMJbRETEMApvERERwyi8ReSvua5LOp1u9jVEPEPhLSIiYhjP7TYX8bJCocDOzg4fHx/09/eTSqWYnp4mHo9zfn4OwPLyMpZlsb+/z/r6On6/n0AggG3bWJZFuVwmm83S1tZGZ2cni4uLALy8vJBOp7m6uiIcDrO2tmbEkg0RE6nzFvGISqXC3t4ejuNQLBYJBoMcHh5yfX1NIpFga2uLWCxGPp/n/f2dubk5VldXKRQKxONxVlZWAJidncW2bTY3NxkZGeHg4ACAy8tLbNvGdV0uLi44PT1tZrkiX5o6bxGPOD4+plarMTU1BcDb2xu3t7eEQiGGhoYAiEajbGxsUK1W6e7upre3F4BYLMb29jYPDw88PT0xMDAAQDKZBL7/845EIgQCAQAsy+L5+fk/VyjiHQpvEY/w+XyMjY0xPz//eVav10kkEp/fjUaDlpaWX8bdP57/bqPyz+9la/OyyL+jsbmIR0SjUUqlEq+vrwA4jsPd3R2Pj4+cnZ0BcHJywuDgIH19fdzf33NzcwPA0dERw8PDdHV1EQqFqFQqAOTzeRzHaU5BIh6mzlvEIyKRCBMTE0xOTtLe3k5PTw+jo6NYloXruiwsLNBoNFhaWsLv95PJZJiZmcHn89HR0UEmkwEgl8uRzWZpbW0lGAySy+XY3d1tcnUi3qJXxUQ8rF6vMz4+TqlUavZVROQPaGwuIiJiGHXeIiIihlHnLSIiYhiFt4iIiGEU3iIiIoZReIuIiBhG4S0iImIYhbeIiIhhvgE8a4svFlhDAQAAAABJRU5ErkJggg==\n",
            "text/plain": [
              "<Figure size 576x396 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "sEL0wOBxbVts",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "submission = pd.read_csv('/content/DataScienceSS20Project/nlp-getting-started/sample_submission.csv')"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "HtGH1XxDar75",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "model.trainable=False\n",
        "\n",
        "optimizer = tf.keras.optimizers.Adam(lr=0.00001)\n",
        "model.compile(loss='binary_crossentropy', optimizer=optimizer, metrics=['accuracy'])\n",
        "model.load_weights('model.h5')\n",
        "train_pred = model.predict(test_input)\n"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "cqa-y_-oQ3Zr",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "submission['target'] = train_pred.round().astype(int)\n",
        "submission.to_csv('submission.csv', index=False)"
      ],
      "execution_count": 0,
      "outputs": []
    }
  ]
}