notebooks/finetuneBERT.ipynb

{
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "view-in-github",
        "colab_type": "text"
      },
      "source": [
        "<a href=\"https://colab.research.google.com/github/mc-cat-tty/PlaceRank/blob/BERTembeddings/finetuneBERT.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "48jg1LDR-E-r"
      },
      "source": [
        "### finetuning BERT for sentiment analysis"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 1,
      "metadata": {
        "id": "4KivTNCL-E-t"
      },
      "outputs": [],
      "source": [
        "import os\n",
        "import re\n",
        "import string\n",
        "from tqdm import tqdm\n",
        "import numpy as np\n",
        "import pandas as pd\n",
        "import matplotlib.pyplot as plt\n",
        "import nltk\n",
        "from nltk.corpus import stopwords\n",
        "from nltk.tokenize import word_tokenize\n",
        "\n",
        "%matplotlib inline"
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "from google.colab import drive\n",
        "drive.mount('/content/drive')"
      ],
      "metadata": {
        "id": "rOgyVa3_Ir02",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "4112d561-cd6c-4e30-d374-0cb34515ad15"
      },
      "execution_count": 2,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Mounted at /content/drive\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 3,
      "metadata": {
        "id": "VT-P7fU7-E-u"
      },
      "outputs": [],
      "source": [
        "file1_path = 'drive/MyDrive/Colab Notebooks/inputs/nyc_listings.csv'\n",
        "file2_path = 'drive/MyDrive/Colab Notebooks/inputs/nyc_reviews.csv'\n",
        "selected1_cols = ['id', 'review_scores_rating']\n",
        "selected2_cols = ['listing_id', 'comments']\n",
        "df1 = pd.read_csv(file1_path, usecols=selected1_cols)\n",
        "df2 = pd.read_csv(file2_path, usecols=selected2_cols)\n",
        "df1.rename(columns={'id': 'listing_id'}, inplace=True)\n",
        "df = pd.merge(df1, df2)\n",
        "df.rename(columns={'listing_id': 'id'}, inplace=True)\n",
        "df.rename(columns={'review_scores_rating': 'rating'}, inplace=True)\n",
        "df.rename(columns={'comments': 'review'}, inplace=True)\n",
        "new_cols = ['id', 'review', 'rating']\n",
        "dfs = df[new_cols]\n",
        "dfs['rating'] = dfs['rating'].round().astype(int)\n",
        "\n",
        "\n",
        "positive = [4, 5]\n",
        "negative = [1, 2, 3]\n",
        "\n",
        "def sentiment(rating):\n",
        "  if rating in positive:\n",
        "    return 1\n",
        "  else:\n",
        "    return 0\n",
        "\n",
        "dfs['target'] = dfs['rating'].apply(sentiment)\n",
        "dfs.rename(columns={'target': 'label'}, inplace=True)\n",
        "\n"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 4,
      "metadata": {
        "id": "pQb6VeE8-E-v"
      },
      "outputs": [],
      "source": [
        "dfs.drop(['rating'], inplace=True, axis=1)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 5,
      "metadata": {
        "id": "9eL5AKOz-E-v",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 206
        },
        "outputId": "8b115a51-eb71-43e0-9e34-287b53487c41"
      },
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "         id                                             review  label\n",
              "0  21935608  Mike was very responsive and accommodating. Th...      1\n",
              "1  21935608  I couldn’t be more satisfied with my stay at M...      1\n",
              "2  21935608  Mike's space was absolutely lovely and in a wo...      1\n",
              "3  21935608  The apartment was just exactly what I was look...      1\n",
              "4  21935608  Mike and Zoe were great hosts. Mike wrote a ve...      1"
            ],
            "text/html": [
              "\n",
              "  <div id=\"df-ec025853-56ba-44ac-a0c9-82fac785109f\" class=\"colab-df-container\">\n",
              "    <div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>id</th>\n",
              "      <th>review</th>\n",
              "      <th>label</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>21935608</td>\n",
              "      <td>Mike was very responsive and accommodating. Th...</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>21935608</td>\n",
              "      <td>I couldn’t be more satisfied with my stay at M...</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>21935608</td>\n",
              "      <td>Mike's space was absolutely lovely and in a wo...</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>21935608</td>\n",
              "      <td>The apartment was just exactly what I was look...</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>21935608</td>\n",
              "      <td>Mike and Zoe were great hosts. Mike wrote a ve...</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>\n",
              "    <div class=\"colab-df-buttons\">\n",
              "\n",
              "  <div class=\"colab-df-container\">\n",
              "    <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-ec025853-56ba-44ac-a0c9-82fac785109f')\"\n",
              "            title=\"Convert this dataframe to an interactive table.\"\n",
              "            style=\"display:none;\">\n",
              "\n",
              "  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\" viewBox=\"0 -960 960 960\">\n",
              "    <path d=\"M120-120v-720h720v720H120Zm60-500h600v-160H180v160Zm220 220h160v-160H400v160Zm0 220h160v-160H400v160ZM180-400h160v-160H180v160Zm440 0h160v-160H620v160ZM180-180h160v-160H180v160Zm440 0h160v-160H620v160Z\"/>\n",
              "  </svg>\n",
              "    </button>\n",
              "\n",
              "  <style>\n",
              "    .colab-df-container {\n",
              "      display:flex;\n",
              "      gap: 12px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert {\n",
              "      background-color: #E8F0FE;\n",
              "      border: none;\n",
              "      border-radius: 50%;\n",
              "      cursor: pointer;\n",
              "      display: none;\n",
              "      fill: #1967D2;\n",
              "      height: 32px;\n",
              "      padding: 0 0 0 0;\n",
              "      width: 32px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert:hover {\n",
              "      background-color: #E2EBFA;\n",
              "      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
              "      fill: #174EA6;\n",
              "    }\n",
              "\n",
              "    .colab-df-buttons div {\n",
              "      margin-bottom: 4px;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert {\n",
              "      background-color: #3B4455;\n",
              "      fill: #D2E3FC;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert:hover {\n",
              "      background-color: #434B5C;\n",
              "      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
              "      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
              "      fill: #FFFFFF;\n",
              "    }\n",
              "  </style>\n",
              "\n",
              "    <script>\n",
              "      const buttonEl =\n",
              "        document.querySelector('#df-ec025853-56ba-44ac-a0c9-82fac785109f button.colab-df-convert');\n",
              "      buttonEl.style.display =\n",
              "        google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "\n",
              "      async function convertToInteractive(key) {\n",
              "        const element = document.querySelector('#df-ec025853-56ba-44ac-a0c9-82fac785109f');\n",
              "        const dataTable =\n",
              "          await google.colab.kernel.invokeFunction('convertToInteractive',\n",
              "                                                    [key], {});\n",
              "        if (!dataTable) return;\n",
              "\n",
              "        const docLinkHtml = 'Like what you see? Visit the ' +\n",
              "          '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n",
              "          + ' to learn more about interactive tables.';\n",
              "        element.innerHTML = '';\n",
              "        dataTable['output_type'] = 'display_data';\n",
              "        await google.colab.output.renderOutput(dataTable, element);\n",
              "        const docLink = document.createElement('div');\n",
              "        docLink.innerHTML = docLinkHtml;\n",
              "        element.appendChild(docLink);\n",
              "      }\n",
              "    </script>\n",
              "  </div>\n",
              "\n",
              "\n",
              "<div id=\"df-e83d2b17-bfe8-45aa-8a9b-0a4220586e9c\">\n",
              "  <button class=\"colab-df-quickchart\" onclick=\"quickchart('df-e83d2b17-bfe8-45aa-8a9b-0a4220586e9c')\"\n",
              "            title=\"Suggest charts\"\n",
              "            style=\"display:none;\">\n",
              "\n",
              "<svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n",
              "     width=\"24px\">\n",
              "    <g>\n",
              "        <path d=\"M19 3H5c-1.1 0-2 .9-2 2v14c0 1.1.9 2 2 2h14c1.1 0 2-.9 2-2V5c0-1.1-.9-2-2-2zM9 17H7v-7h2v7zm4 0h-2V7h2v10zm4 0h-2v-4h2v4z\"/>\n",
              "    </g>\n",
              "</svg>\n",
              "  </button>\n",
              "\n",
              "<style>\n",
              "  .colab-df-quickchart {\n",
              "      --bg-color: #E8F0FE;\n",
              "      --fill-color: #1967D2;\n",
              "      --hover-bg-color: #E2EBFA;\n",
              "      --hover-fill-color: #174EA6;\n",
              "      --disabled-fill-color: #AAA;\n",
              "      --disabled-bg-color: #DDD;\n",
              "  }\n",
              "\n",
              "  [theme=dark] .colab-df-quickchart {\n",
              "      --bg-color: #3B4455;\n",
              "      --fill-color: #D2E3FC;\n",
              "      --hover-bg-color: #434B5C;\n",
              "      --hover-fill-color: #FFFFFF;\n",
              "      --disabled-bg-color: #3B4455;\n",
              "      --disabled-fill-color: #666;\n",
              "  }\n",
              "\n",
              "  .colab-df-quickchart {\n",
              "    background-color: var(--bg-color);\n",
              "    border: none;\n",
              "    border-radius: 50%;\n",
              "    cursor: pointer;\n",
              "    display: none;\n",
              "    fill: var(--fill-color);\n",
              "    height: 32px;\n",
              "    padding: 0;\n",
              "    width: 32px;\n",
              "  }\n",
              "\n",
              "  .colab-df-quickchart:hover {\n",
              "    background-color: var(--hover-bg-color);\n",
              "    box-shadow: 0 1px 2px rgba(60, 64, 67, 0.3), 0 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
              "    fill: var(--button-hover-fill-color);\n",
              "  }\n",
              "\n",
              "  .colab-df-quickchart-complete:disabled,\n",
              "  .colab-df-quickchart-complete:disabled:hover {\n",
              "    background-color: var(--disabled-bg-color);\n",
              "    fill: var(--disabled-fill-color);\n",
              "    box-shadow: none;\n",
              "  }\n",
              "\n",
              "  .colab-df-spinner {\n",
              "    border: 2px solid var(--fill-color);\n",
              "    border-color: transparent;\n",
              "    border-bottom-color: var(--fill-color);\n",
              "    animation:\n",
              "      spin 1s steps(1) infinite;\n",
              "  }\n",
              "\n",
              "  @keyframes spin {\n",
              "    0% {\n",
              "      border-color: transparent;\n",
              "      border-bottom-color: var(--fill-color);\n",
              "      border-left-color: var(--fill-color);\n",
              "    }\n",
              "    20% {\n",
              "      border-color: transparent;\n",
              "      border-left-color: var(--fill-color);\n",
              "      border-top-color: var(--fill-color);\n",
              "    }\n",
              "    30% {\n",
              "      border-color: transparent;\n",
              "      border-left-color: var(--fill-color);\n",
              "      border-top-color: var(--fill-color);\n",
              "      border-right-color: var(--fill-color);\n",
              "    }\n",
              "    40% {\n",
              "      border-color: transparent;\n",
              "      border-right-color: var(--fill-color);\n",
              "      border-top-color: var(--fill-color);\n",
              "    }\n",
              "    60% {\n",
              "      border-color: transparent;\n",
              "      border-right-color: var(--fill-color);\n",
              "    }\n",
              "    80% {\n",
              "      border-color: transparent;\n",
              "      border-right-color: var(--fill-color);\n",
              "      border-bottom-color: var(--fill-color);\n",
              "    }\n",
              "    90% {\n",
              "      border-color: transparent;\n",
              "      border-bottom-color: var(--fill-color);\n",
              "    }\n",
              "  }\n",
              "</style>\n",
              "\n",
              "  <script>\n",
              "    async function quickchart(key) {\n",
              "      const quickchartButtonEl =\n",
              "        document.querySelector('#' + key + ' button');\n",
              "      quickchartButtonEl.disabled = true;  // To prevent multiple clicks.\n",
              "      quickchartButtonEl.classList.add('colab-df-spinner');\n",
              "      try {\n",
              "        const charts = await google.colab.kernel.invokeFunction(\n",
              "            'suggestCharts', [key], {});\n",
              "      } catch (error) {\n",
              "        console.error('Error during call to suggestCharts:', error);\n",
              "      }\n",
              "      quickchartButtonEl.classList.remove('colab-df-spinner');\n",
              "      quickchartButtonEl.classList.add('colab-df-quickchart-complete');\n",
              "    }\n",
              "    (() => {\n",
              "      let quickchartButtonEl =\n",
              "        document.querySelector('#df-e83d2b17-bfe8-45aa-8a9b-0a4220586e9c button');\n",
              "      quickchartButtonEl.style.display =\n",
              "        google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "    })();\n",
              "  </script>\n",
              "</div>\n",
              "    </div>\n",
              "  </div>\n"
            ]
          },
          "metadata": {},
          "execution_count": 5
        }
      ],
      "source": [
        "dfs.head()"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 6,
      "metadata": {
        "id": "yxzxPw9V-E-v",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "577adfac-6819-4e1c-cdf5-b48a1e99e659"
      },
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "1    999846\n",
              "0      1449\n",
              "Name: label, dtype: int64"
            ]
          },
          "metadata": {},
          "execution_count": 6
        }
      ],
      "source": [
        "balance_counts = dfs.groupby('label')['label'].agg('count').values\n",
        "dfs['label'].value_counts()"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 7,
      "metadata": {
        "id": "HSfyiq2H-E-w"
      },
      "outputs": [],
      "source": [
        "dfs.review = dfs.review.astype(str)"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "mMx9Yidq-E-w"
      },
      "source": [
        "### building a test dataset"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 8,
      "metadata": {
        "id": "dFxSKtU3-E-w",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "2e274657-53bd-4349-b05a-ae85594d1361"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "<ipython-input-8-6e2116672c47>:5: SettingWithCopyWarning: \n",
            "A value is trying to be set on a copy of a slice from a DataFrame\n",
            "\n",
            "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
            "  neg_dfs.drop(['label'], inplace=True, axis=1)\n"
          ]
        }
      ],
      "source": [
        "# negative sentiment\n",
        "neg_dfs = dfs[(dfs[\"label\"] == 0)]\n",
        "\n",
        "# drop the label\n",
        "neg_dfs.drop(['label'], inplace=True, axis=1)\n",
        "\n",
        "new_dfs = dfs[(dfs[\"label\"] == 1)].sample(300)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 15,
      "metadata": {
        "id": "lV5Erkbp-E-w",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "0d8672b9-d5eb-4f08-c034-a13f16a6c900"
      },
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "id                   review                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                 \n",
              "34487108             .                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                          2\n",
              "302280               \"Tammy's place is very well situated in Harlem. It's close to shops, restaurants and subway.<br/>However, we were 6 people and we paid an expensive price for just one bedroom and two couches. The kitchen isn't well equipped to cook (oven handle was broken, stoves are in a very bad shape...) and there is only 4 chairs.<br/>The apartment wasn't so clean, the bed sheet were dirty and we could hear/feel the subway under us.\"                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                   1\n",
              "625924887219854517   The only thing I will say about this space is that it is a 5th floor walk up. The space was clean.                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                         1\n",
              "639038485645896375   terrible cancellation policy, terrible building. great neighborhood though                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                 1\n",
              "629622149018334971   This place is pretty small and cheap if you are okay with just managing. There is garbage dirty smell every time you walk out out the building (lot of garbage in the neighborhood  and prone to rats). Communication is bad so there is not much you can ask/request from the host. There is no lock to your room door nor the apartment I stayed in so I did not feel very safe with my things when I am outside, there was only one door locked which was the main door of the five floor building. The bed was also very dirty and did not seemed to be washed recently, I had to put my own mattress and sheets to avid bugs and ticks.                                                                                                                                                                                                                                                                                                                                                                                               1\n",
              "                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                               ..\n",
              "39729023             Fake host! Ad was by a man and his (Fake friendly) girlfriend  showed up.  The apartment  is cold at night and she only  gave  me one blanket , dusty, people in the  building  were nasty to  me and my daughter(because  we were black) This  place is  not wori the stay at all. Especially  when  the  host is rude, she made my  daughter  feel  so bad.😶😣                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                            1\n",
              "39558009             This is what I felt during stay<br/>1.Geographical drawbacks(trains after midnight)<br/>2. They did not informed that We were gonna live with one of their family.<br/>3. One of the woman smoked weed in the living room<br/>4. A few lights. Hard to see in the night<br/>5. Bath room near bed room has too much smokes coming from vent.<br/>6. No air circulation in the house.                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                       1\n",
              "39556412             I'll start with the positives. The apartment is clean and bright, and located in a quiet, lovely neighborhood near Carl Schurz park and lots of great restaurants. Comfortable bed, and good pressure in the shower. A management company takes care of the details, so communication is relatively good. Now for the downsides. The apartment itself is not very quiet. The walls are so thin you can hear people in the neighboring apartments, both from the living room and the bedroom. One could remedy this issue by being out in NYC all day and using ear plugs at night. Another issue was that \"hotel-style amenities\" were offered in the listing but not provided. This was Rishi's first time hosting, so it's possible he forgot them. But he really should have cleared his stuff out of the closets. There was no space to store our clothes and luggage, and no hooks for coats or other items. Ultimately our biggest issue with the apartment was building security. Next time we will book a place with a doorman.    1\n",
              "39537003             Leïla was super nice! Our flight arrived early and she let us check in earlier than scheduled and we had a late flight and she let us keep our bags at the apartment so that we could continue to enjoy the city! SHE WAS GREAT!                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                           1\n",
              "1039133098605886642  Cannot write a review. I was not able to stay due to my cat being injured.                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                 1\n",
              "Length: 1448, dtype: int64"
            ]
          },
          "metadata": {},
          "execution_count": 15
        }
      ],
      "source": [
        "neg_dfs.value_counts()"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 16,
      "metadata": {
        "id": "LnASp28X-E-w"
      },
      "outputs": [],
      "source": [
        "new_dfs.drop(['label'], inplace=True, axis=1)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 17,
      "metadata": {
        "id": "-eQOxxjQ-E-w",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 363
        },
        "outputId": "f190517a-8f8a-4bbd-e9c9-889b724751d4"
      },
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "                        id                                             review\n",
              "409119  950316243638576156  Had an absolute perfect stay at Jacqueline’s b...\n",
              "822444            21714930                                               Tuff\n",
              "26476               749276  Wilfredo was a great host, fast response, and ...\n",
              "554999            30264512  It was cold! I actually left before check out!...\n",
              "544305            45359389   Great place, lots of space, and thoughtful hosts\n",
              "121609            15665152  A lovely retreat after busy days touring Manha...\n",
              "275975            35545434  Location is great, and it's a clean, affordabl...\n",
              "766883  974559250469050993  Friendly host, clean room, great amenities in ...\n",
              "770056  686302065098279368          Great location! Spacious and pretty room.\n",
              "323100  839929647945498049  I had such a great stay! Olivia was very hospi..."
            ],
            "text/html": [
              "\n",
              "  <div id=\"df-55f089ac-be69-45a0-a8be-5fe8a1322b81\" class=\"colab-df-container\">\n",
              "    <div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>id</th>\n",
              "      <th>review</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>409119</th>\n",
              "      <td>950316243638576156</td>\n",
              "      <td>Had an absolute perfect stay at Jacqueline’s b...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>822444</th>\n",
              "      <td>21714930</td>\n",
              "      <td>Tuff</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>26476</th>\n",
              "      <td>749276</td>\n",
              "      <td>Wilfredo was a great host, fast response, and ...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>554999</th>\n",
              "      <td>30264512</td>\n",
              "      <td>It was cold! I actually left before check out!...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>544305</th>\n",
              "      <td>45359389</td>\n",
              "      <td>Great place, lots of space, and thoughtful hosts</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>121609</th>\n",
              "      <td>15665152</td>\n",
              "      <td>A lovely retreat after busy days touring Manha...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>275975</th>\n",
              "      <td>35545434</td>\n",
              "      <td>Location is great, and it's a clean, affordabl...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>766883</th>\n",
              "      <td>974559250469050993</td>\n",
              "      <td>Friendly host, clean room, great amenities in ...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>770056</th>\n",
              "      <td>686302065098279368</td>\n",
              "      <td>Great location! Spacious and pretty room.</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>323100</th>\n",
              "      <td>839929647945498049</td>\n",
              "      <td>I had such a great stay! Olivia was very hospi...</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>\n",
              "    <div class=\"colab-df-buttons\">\n",
              "\n",
              "  <div class=\"colab-df-container\">\n",
              "    <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-55f089ac-be69-45a0-a8be-5fe8a1322b81')\"\n",
              "            title=\"Convert this dataframe to an interactive table.\"\n",
              "            style=\"display:none;\">\n",
              "\n",
              "  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\" viewBox=\"0 -960 960 960\">\n",
              "    <path d=\"M120-120v-720h720v720H120Zm60-500h600v-160H180v160Zm220 220h160v-160H400v160Zm0 220h160v-160H400v160ZM180-400h160v-160H180v160Zm440 0h160v-160H620v160ZM180-180h160v-160H180v160Zm440 0h160v-160H620v160Z\"/>\n",
              "  </svg>\n",
              "    </button>\n",
              "\n",
              "  <style>\n",
              "    .colab-df-container {\n",
              "      display:flex;\n",
              "      gap: 12px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert {\n",
              "      background-color: #E8F0FE;\n",
              "      border: none;\n",
              "      border-radius: 50%;\n",
              "      cursor: pointer;\n",
              "      display: none;\n",
              "      fill: #1967D2;\n",
              "      height: 32px;\n",
              "      padding: 0 0 0 0;\n",
              "      width: 32px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert:hover {\n",
              "      background-color: #E2EBFA;\n",
              "      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
              "      fill: #174EA6;\n",
              "    }\n",
              "\n",
              "    .colab-df-buttons div {\n",
              "      margin-bottom: 4px;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert {\n",
              "      background-color: #3B4455;\n",
              "      fill: #D2E3FC;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert:hover {\n",
              "      background-color: #434B5C;\n",
              "      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
              "      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
              "      fill: #FFFFFF;\n",
              "    }\n",
              "  </style>\n",
              "\n",
              "    <script>\n",
              "      const buttonEl =\n",
              "        document.querySelector('#df-55f089ac-be69-45a0-a8be-5fe8a1322b81 button.colab-df-convert');\n",
              "      buttonEl.style.display =\n",
              "        google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "\n",
              "      async function convertToInteractive(key) {\n",
              "        const element = document.querySelector('#df-55f089ac-be69-45a0-a8be-5fe8a1322b81');\n",
              "        const dataTable =\n",
              "          await google.colab.kernel.invokeFunction('convertToInteractive',\n",
              "                                                    [key], {});\n",
              "        if (!dataTable) return;\n",
              "\n",
              "        const docLinkHtml = 'Like what you see? Visit the ' +\n",
              "          '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n",
              "          + ' to learn more about interactive tables.';\n",
              "        element.innerHTML = '';\n",
              "        dataTable['output_type'] = 'display_data';\n",
              "        await google.colab.output.renderOutput(dataTable, element);\n",
              "        const docLink = document.createElement('div');\n",
              "        docLink.innerHTML = docLinkHtml;\n",
              "        element.appendChild(docLink);\n",
              "      }\n",
              "    </script>\n",
              "  </div>\n",
              "\n",
              "\n",
              "<div id=\"df-96d876f7-653f-4d7f-9c03-5757682bc8e2\">\n",
              "  <button class=\"colab-df-quickchart\" onclick=\"quickchart('df-96d876f7-653f-4d7f-9c03-5757682bc8e2')\"\n",
              "            title=\"Suggest charts\"\n",
              "            style=\"display:none;\">\n",
              "\n",
              "<svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n",
              "     width=\"24px\">\n",
              "    <g>\n",
              "        <path d=\"M19 3H5c-1.1 0-2 .9-2 2v14c0 1.1.9 2 2 2h14c1.1 0 2-.9 2-2V5c0-1.1-.9-2-2-2zM9 17H7v-7h2v7zm4 0h-2V7h2v10zm4 0h-2v-4h2v4z\"/>\n",
              "    </g>\n",
              "</svg>\n",
              "  </button>\n",
              "\n",
              "<style>\n",
              "  .colab-df-quickchart {\n",
              "      --bg-color: #E8F0FE;\n",
              "      --fill-color: #1967D2;\n",
              "      --hover-bg-color: #E2EBFA;\n",
              "      --hover-fill-color: #174EA6;\n",
              "      --disabled-fill-color: #AAA;\n",
              "      --disabled-bg-color: #DDD;\n",
              "  }\n",
              "\n",
              "  [theme=dark] .colab-df-quickchart {\n",
              "      --bg-color: #3B4455;\n",
              "      --fill-color: #D2E3FC;\n",
              "      --hover-bg-color: #434B5C;\n",
              "      --hover-fill-color: #FFFFFF;\n",
              "      --disabled-bg-color: #3B4455;\n",
              "      --disabled-fill-color: #666;\n",
              "  }\n",
              "\n",
              "  .colab-df-quickchart {\n",
              "    background-color: var(--bg-color);\n",
              "    border: none;\n",
              "    border-radius: 50%;\n",
              "    cursor: pointer;\n",
              "    display: none;\n",
              "    fill: var(--fill-color);\n",
              "    height: 32px;\n",
              "    padding: 0;\n",
              "    width: 32px;\n",
              "  }\n",
              "\n",
              "  .colab-df-quickchart:hover {\n",
              "    background-color: var(--hover-bg-color);\n",
              "    box-shadow: 0 1px 2px rgba(60, 64, 67, 0.3), 0 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
              "    fill: var(--button-hover-fill-color);\n",
              "  }\n",
              "\n",
              "  .colab-df-quickchart-complete:disabled,\n",
              "  .colab-df-quickchart-complete:disabled:hover {\n",
              "    background-color: var(--disabled-bg-color);\n",
              "    fill: var(--disabled-fill-color);\n",
              "    box-shadow: none;\n",
              "  }\n",
              "\n",
              "  .colab-df-spinner {\n",
              "    border: 2px solid var(--fill-color);\n",
              "    border-color: transparent;\n",
              "    border-bottom-color: var(--fill-color);\n",
              "    animation:\n",
              "      spin 1s steps(1) infinite;\n",
              "  }\n",
              "\n",
              "  @keyframes spin {\n",
              "    0% {\n",
              "      border-color: transparent;\n",
              "      border-bottom-color: var(--fill-color);\n",
              "      border-left-color: var(--fill-color);\n",
              "    }\n",
              "    20% {\n",
              "      border-color: transparent;\n",
              "      border-left-color: var(--fill-color);\n",
              "      border-top-color: var(--fill-color);\n",
              "    }\n",
              "    30% {\n",
              "      border-color: transparent;\n",
              "      border-left-color: var(--fill-color);\n",
              "      border-top-color: var(--fill-color);\n",
              "      border-right-color: var(--fill-color);\n",
              "    }\n",
              "    40% {\n",
              "      border-color: transparent;\n",
              "      border-right-color: var(--fill-color);\n",
              "      border-top-color: var(--fill-color);\n",
              "    }\n",
              "    60% {\n",
              "      border-color: transparent;\n",
              "      border-right-color: var(--fill-color);\n",
              "    }\n",
              "    80% {\n",
              "      border-color: transparent;\n",
              "      border-right-color: var(--fill-color);\n",
              "      border-bottom-color: var(--fill-color);\n",
              "    }\n",
              "    90% {\n",
              "      border-color: transparent;\n",
              "      border-bottom-color: var(--fill-color);\n",
              "    }\n",
              "  }\n",
              "</style>\n",
              "\n",
              "  <script>\n",
              "    async function quickchart(key) {\n",
              "      const quickchartButtonEl =\n",
              "        document.querySelector('#' + key + ' button');\n",
              "      quickchartButtonEl.disabled = true;  // To prevent multiple clicks.\n",
              "      quickchartButtonEl.classList.add('colab-df-spinner');\n",
              "      try {\n",
              "        const charts = await google.colab.kernel.invokeFunction(\n",
              "            'suggestCharts', [key], {});\n",
              "      } catch (error) {\n",
              "        console.error('Error during call to suggestCharts:', error);\n",
              "      }\n",
              "      quickchartButtonEl.classList.remove('colab-df-spinner');\n",
              "      quickchartButtonEl.classList.add('colab-df-quickchart-complete');\n",
              "    }\n",
              "    (() => {\n",
              "      let quickchartButtonEl =\n",
              "        document.querySelector('#df-96d876f7-653f-4d7f-9c03-5757682bc8e2 button');\n",
              "      quickchartButtonEl.style.display =\n",
              "        google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "    })();\n",
              "  </script>\n",
              "</div>\n",
              "    </div>\n",
              "  </div>\n"
            ]
          },
          "metadata": {},
          "execution_count": 17
        }
      ],
      "source": [
        "new_dfs.sample(10)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 18,
      "metadata": {
        "id": "WeWZQNrE-E-x",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 206
        },
        "outputId": "c4f46d77-3cfc-42a9-80e3-7a11b326dd22"
      },
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "           id                                             review\n",
              "321  19531174  Most horrible Airbnb experience I ever had, we...\n",
              "363  22306777  This location is ideal for fans of Brooklyn be...\n",
              "364  22306777  Bartolomeo is a nice and friendly host, the ap...\n",
              "454  28224871  I needed a place near NY Presbyterian Hospital...\n",
              "455  28224871  Unfortunately the host did not provide keys to..."
            ],
            "text/html": [
              "\n",
              "  <div id=\"df-9150c80f-b08b-43f7-8896-c1a8dcc4ad92\" class=\"colab-df-container\">\n",
              "    <div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>id</th>\n",
              "      <th>review</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>321</th>\n",
              "      <td>19531174</td>\n",
              "      <td>Most horrible Airbnb experience I ever had, we...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>363</th>\n",
              "      <td>22306777</td>\n",
              "      <td>This location is ideal for fans of Brooklyn be...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>364</th>\n",
              "      <td>22306777</td>\n",
              "      <td>Bartolomeo is a nice and friendly host, the ap...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>454</th>\n",
              "      <td>28224871</td>\n",
              "      <td>I needed a place near NY Presbyterian Hospital...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>455</th>\n",
              "      <td>28224871</td>\n",
              "      <td>Unfortunately the host did not provide keys to...</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>\n",
              "    <div class=\"colab-df-buttons\">\n",
              "\n",
              "  <div class=\"colab-df-container\">\n",
              "    <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-9150c80f-b08b-43f7-8896-c1a8dcc4ad92')\"\n",
              "            title=\"Convert this dataframe to an interactive table.\"\n",
              "            style=\"display:none;\">\n",
              "\n",
              "  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\" viewBox=\"0 -960 960 960\">\n",
              "    <path d=\"M120-120v-720h720v720H120Zm60-500h600v-160H180v160Zm220 220h160v-160H400v160Zm0 220h160v-160H400v160ZM180-400h160v-160H180v160Zm440 0h160v-160H620v160ZM180-180h160v-160H180v160Zm440 0h160v-160H620v160Z\"/>\n",
              "  </svg>\n",
              "    </button>\n",
              "\n",
              "  <style>\n",
              "    .colab-df-container {\n",
              "      display:flex;\n",
              "      gap: 12px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert {\n",
              "      background-color: #E8F0FE;\n",
              "      border: none;\n",
              "      border-radius: 50%;\n",
              "      cursor: pointer;\n",
              "      display: none;\n",
              "      fill: #1967D2;\n",
              "      height: 32px;\n",
              "      padding: 0 0 0 0;\n",
              "      width: 32px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert:hover {\n",
              "      background-color: #E2EBFA;\n",
              "      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
              "      fill: #174EA6;\n",
              "    }\n",
              "\n",
              "    .colab-df-buttons div {\n",
              "      margin-bottom: 4px;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert {\n",
              "      background-color: #3B4455;\n",
              "      fill: #D2E3FC;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert:hover {\n",
              "      background-color: #434B5C;\n",
              "      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
              "      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
              "      fill: #FFFFFF;\n",
              "    }\n",
              "  </style>\n",
              "\n",
              "    <script>\n",
              "      const buttonEl =\n",
              "        document.querySelector('#df-9150c80f-b08b-43f7-8896-c1a8dcc4ad92 button.colab-df-convert');\n",
              "      buttonEl.style.display =\n",
              "        google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "\n",
              "      async function convertToInteractive(key) {\n",
              "        const element = document.querySelector('#df-9150c80f-b08b-43f7-8896-c1a8dcc4ad92');\n",
              "        const dataTable =\n",
              "          await google.colab.kernel.invokeFunction('convertToInteractive',\n",
              "                                                    [key], {});\n",
              "        if (!dataTable) return;\n",
              "\n",
              "        const docLinkHtml = 'Like what you see? Visit the ' +\n",
              "          '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n",
              "          + ' to learn more about interactive tables.';\n",
              "        element.innerHTML = '';\n",
              "        dataTable['output_type'] = 'display_data';\n",
              "        await google.colab.output.renderOutput(dataTable, element);\n",
              "        const docLink = document.createElement('div');\n",
              "        docLink.innerHTML = docLinkHtml;\n",
              "        element.appendChild(docLink);\n",
              "      }\n",
              "    </script>\n",
              "  </div>\n",
              "\n",
              "\n",
              "<div id=\"df-4989c73d-8cb5-4c68-bb5e-c7d38979e5f0\">\n",
              "  <button class=\"colab-df-quickchart\" onclick=\"quickchart('df-4989c73d-8cb5-4c68-bb5e-c7d38979e5f0')\"\n",
              "            title=\"Suggest charts\"\n",
              "            style=\"display:none;\">\n",
              "\n",
              "<svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n",
              "     width=\"24px\">\n",
              "    <g>\n",
              "        <path d=\"M19 3H5c-1.1 0-2 .9-2 2v14c0 1.1.9 2 2 2h14c1.1 0 2-.9 2-2V5c0-1.1-.9-2-2-2zM9 17H7v-7h2v7zm4 0h-2V7h2v10zm4 0h-2v-4h2v4z\"/>\n",
              "    </g>\n",
              "</svg>\n",
              "  </button>\n",
              "\n",
              "<style>\n",
              "  .colab-df-quickchart {\n",
              "      --bg-color: #E8F0FE;\n",
              "      --fill-color: #1967D2;\n",
              "      --hover-bg-color: #E2EBFA;\n",
              "      --hover-fill-color: #174EA6;\n",
              "      --disabled-fill-color: #AAA;\n",
              "      --disabled-bg-color: #DDD;\n",
              "  }\n",
              "\n",
              "  [theme=dark] .colab-df-quickchart {\n",
              "      --bg-color: #3B4455;\n",
              "      --fill-color: #D2E3FC;\n",
              "      --hover-bg-color: #434B5C;\n",
              "      --hover-fill-color: #FFFFFF;\n",
              "      --disabled-bg-color: #3B4455;\n",
              "      --disabled-fill-color: #666;\n",
              "  }\n",
              "\n",
              "  .colab-df-quickchart {\n",
              "    background-color: var(--bg-color);\n",
              "    border: none;\n",
              "    border-radius: 50%;\n",
              "    cursor: pointer;\n",
              "    display: none;\n",
              "    fill: var(--fill-color);\n",
              "    height: 32px;\n",
              "    padding: 0;\n",
              "    width: 32px;\n",
              "  }\n",
              "\n",
              "  .colab-df-quickchart:hover {\n",
              "    background-color: var(--hover-bg-color);\n",
              "    box-shadow: 0 1px 2px rgba(60, 64, 67, 0.3), 0 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
              "    fill: var(--button-hover-fill-color);\n",
              "  }\n",
              "\n",
              "  .colab-df-quickchart-complete:disabled,\n",
              "  .colab-df-quickchart-complete:disabled:hover {\n",
              "    background-color: var(--disabled-bg-color);\n",
              "    fill: var(--disabled-fill-color);\n",
              "    box-shadow: none;\n",
              "  }\n",
              "\n",
              "  .colab-df-spinner {\n",
              "    border: 2px solid var(--fill-color);\n",
              "    border-color: transparent;\n",
              "    border-bottom-color: var(--fill-color);\n",
              "    animation:\n",
              "      spin 1s steps(1) infinite;\n",
              "  }\n",
              "\n",
              "  @keyframes spin {\n",
              "    0% {\n",
              "      border-color: transparent;\n",
              "      border-bottom-color: var(--fill-color);\n",
              "      border-left-color: var(--fill-color);\n",
              "    }\n",
              "    20% {\n",
              "      border-color: transparent;\n",
              "      border-left-color: var(--fill-color);\n",
              "      border-top-color: var(--fill-color);\n",
              "    }\n",
              "    30% {\n",
              "      border-color: transparent;\n",
              "      border-left-color: var(--fill-color);\n",
              "      border-top-color: var(--fill-color);\n",
              "      border-right-color: var(--fill-color);\n",
              "    }\n",
              "    40% {\n",
              "      border-color: transparent;\n",
              "      border-right-color: var(--fill-color);\n",
              "      border-top-color: var(--fill-color);\n",
              "    }\n",
              "    60% {\n",
              "      border-color: transparent;\n",
              "      border-right-color: var(--fill-color);\n",
              "    }\n",
              "    80% {\n",
              "      border-color: transparent;\n",
              "      border-right-color: var(--fill-color);\n",
              "      border-bottom-color: var(--fill-color);\n",
              "    }\n",
              "    90% {\n",
              "      border-color: transparent;\n",
              "      border-bottom-color: var(--fill-color);\n",
              "    }\n",
              "  }\n",
              "</style>\n",
              "\n",
              "  <script>\n",
              "    async function quickchart(key) {\n",
              "      const quickchartButtonEl =\n",
              "        document.querySelector('#' + key + ' button');\n",
              "      quickchartButtonEl.disabled = true;  // To prevent multiple clicks.\n",
              "      quickchartButtonEl.classList.add('colab-df-spinner');\n",
              "      try {\n",
              "        const charts = await google.colab.kernel.invokeFunction(\n",
              "            'suggestCharts', [key], {});\n",
              "      } catch (error) {\n",
              "        console.error('Error during call to suggestCharts:', error);\n",
              "      }\n",
              "      quickchartButtonEl.classList.remove('colab-df-spinner');\n",
              "      quickchartButtonEl.classList.add('colab-df-quickchart-complete');\n",
              "    }\n",
              "    (() => {\n",
              "      let quickchartButtonEl =\n",
              "        document.querySelector('#df-4989c73d-8cb5-4c68-bb5e-c7d38979e5f0 button');\n",
              "      quickchartButtonEl.style.display =\n",
              "        google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "    })();\n",
              "  </script>\n",
              "</div>\n",
              "    </div>\n",
              "  </div>\n"
            ]
          },
          "metadata": {},
          "execution_count": 18
        }
      ],
      "source": [
        "neg_dfs.head()"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 19,
      "metadata": {
        "id": "4KmeUyV--E-x",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 206
        },
        "outputId": "3bec19ca-7c48-403a-9153-42d94b7b23c9"
      },
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "                        id                                             review\n",
              "411287            30557920                                         Great stay\n",
              "647158            28901091  Wonderful spot to stay for exploring NYC! Woul...\n",
              "979536  649819721923399503                               Thank you, Shamaine!\n",
              "939842            40797714                             Everything was awesome\n",
              "302012             4059620  Excellent location, great communication from h..."
            ],
            "text/html": [
              "\n",
              "  <div id=\"df-1f86843b-0628-4702-a1d1-82919f001c0b\" class=\"colab-df-container\">\n",
              "    <div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>id</th>\n",
              "      <th>review</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>411287</th>\n",
              "      <td>30557920</td>\n",
              "      <td>Great stay</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>647158</th>\n",
              "      <td>28901091</td>\n",
              "      <td>Wonderful spot to stay for exploring NYC! Woul...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>979536</th>\n",
              "      <td>649819721923399503</td>\n",
              "      <td>Thank you, Shamaine!</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>939842</th>\n",
              "      <td>40797714</td>\n",
              "      <td>Everything was awesome</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>302012</th>\n",
              "      <td>4059620</td>\n",
              "      <td>Excellent location, great communication from h...</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>\n",
              "    <div class=\"colab-df-buttons\">\n",
              "\n",
              "  <div class=\"colab-df-container\">\n",
              "    <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-1f86843b-0628-4702-a1d1-82919f001c0b')\"\n",
              "            title=\"Convert this dataframe to an interactive table.\"\n",
              "            style=\"display:none;\">\n",
              "\n",
              "  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\" viewBox=\"0 -960 960 960\">\n",
              "    <path d=\"M120-120v-720h720v720H120Zm60-500h600v-160H180v160Zm220 220h160v-160H400v160Zm0 220h160v-160H400v160ZM180-400h160v-160H180v160Zm440 0h160v-160H620v160ZM180-180h160v-160H180v160Zm440 0h160v-160H620v160Z\"/>\n",
              "  </svg>\n",
              "    </button>\n",
              "\n",
              "  <style>\n",
              "    .colab-df-container {\n",
              "      display:flex;\n",
              "      gap: 12px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert {\n",
              "      background-color: #E8F0FE;\n",
              "      border: none;\n",
              "      border-radius: 50%;\n",
              "      cursor: pointer;\n",
              "      display: none;\n",
              "      fill: #1967D2;\n",
              "      height: 32px;\n",
              "      padding: 0 0 0 0;\n",
              "      width: 32px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert:hover {\n",
              "      background-color: #E2EBFA;\n",
              "      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
              "      fill: #174EA6;\n",
              "    }\n",
              "\n",
              "    .colab-df-buttons div {\n",
              "      margin-bottom: 4px;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert {\n",
              "      background-color: #3B4455;\n",
              "      fill: #D2E3FC;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert:hover {\n",
              "      background-color: #434B5C;\n",
              "      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
              "      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
              "      fill: #FFFFFF;\n",
              "    }\n",
              "  </style>\n",
              "\n",
              "    <script>\n",
              "      const buttonEl =\n",
              "        document.querySelector('#df-1f86843b-0628-4702-a1d1-82919f001c0b button.colab-df-convert');\n",
              "      buttonEl.style.display =\n",
              "        google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "\n",
              "      async function convertToInteractive(key) {\n",
              "        const element = document.querySelector('#df-1f86843b-0628-4702-a1d1-82919f001c0b');\n",
              "        const dataTable =\n",
              "          await google.colab.kernel.invokeFunction('convertToInteractive',\n",
              "                                                    [key], {});\n",
              "        if (!dataTable) return;\n",
              "\n",
              "        const docLinkHtml = 'Like what you see? Visit the ' +\n",
              "          '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n",
              "          + ' to learn more about interactive tables.';\n",
              "        element.innerHTML = '';\n",
              "        dataTable['output_type'] = 'display_data';\n",
              "        await google.colab.output.renderOutput(dataTable, element);\n",
              "        const docLink = document.createElement('div');\n",
              "        docLink.innerHTML = docLinkHtml;\n",
              "        element.appendChild(docLink);\n",
              "      }\n",
              "    </script>\n",
              "  </div>\n",
              "\n",
              "\n",
              "<div id=\"df-a6f312d0-4d04-4b75-96d1-f065fe203119\">\n",
              "  <button class=\"colab-df-quickchart\" onclick=\"quickchart('df-a6f312d0-4d04-4b75-96d1-f065fe203119')\"\n",
              "            title=\"Suggest charts\"\n",
              "            style=\"display:none;\">\n",
              "\n",
              "<svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n",
              "     width=\"24px\">\n",
              "    <g>\n",
              "        <path d=\"M19 3H5c-1.1 0-2 .9-2 2v14c0 1.1.9 2 2 2h14c1.1 0 2-.9 2-2V5c0-1.1-.9-2-2-2zM9 17H7v-7h2v7zm4 0h-2V7h2v10zm4 0h-2v-4h2v4z\"/>\n",
              "    </g>\n",
              "</svg>\n",
              "  </button>\n",
              "\n",
              "<style>\n",
              "  .colab-df-quickchart {\n",
              "      --bg-color: #E8F0FE;\n",
              "      --fill-color: #1967D2;\n",
              "      --hover-bg-color: #E2EBFA;\n",
              "      --hover-fill-color: #174EA6;\n",
              "      --disabled-fill-color: #AAA;\n",
              "      --disabled-bg-color: #DDD;\n",
              "  }\n",
              "\n",
              "  [theme=dark] .colab-df-quickchart {\n",
              "      --bg-color: #3B4455;\n",
              "      --fill-color: #D2E3FC;\n",
              "      --hover-bg-color: #434B5C;\n",
              "      --hover-fill-color: #FFFFFF;\n",
              "      --disabled-bg-color: #3B4455;\n",
              "      --disabled-fill-color: #666;\n",
              "  }\n",
              "\n",
              "  .colab-df-quickchart {\n",
              "    background-color: var(--bg-color);\n",
              "    border: none;\n",
              "    border-radius: 50%;\n",
              "    cursor: pointer;\n",
              "    display: none;\n",
              "    fill: var(--fill-color);\n",
              "    height: 32px;\n",
              "    padding: 0;\n",
              "    width: 32px;\n",
              "  }\n",
              "\n",
              "  .colab-df-quickchart:hover {\n",
              "    background-color: var(--hover-bg-color);\n",
              "    box-shadow: 0 1px 2px rgba(60, 64, 67, 0.3), 0 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
              "    fill: var(--button-hover-fill-color);\n",
              "  }\n",
              "\n",
              "  .colab-df-quickchart-complete:disabled,\n",
              "  .colab-df-quickchart-complete:disabled:hover {\n",
              "    background-color: var(--disabled-bg-color);\n",
              "    fill: var(--disabled-fill-color);\n",
              "    box-shadow: none;\n",
              "  }\n",
              "\n",
              "  .colab-df-spinner {\n",
              "    border: 2px solid var(--fill-color);\n",
              "    border-color: transparent;\n",
              "    border-bottom-color: var(--fill-color);\n",
              "    animation:\n",
              "      spin 1s steps(1) infinite;\n",
              "  }\n",
              "\n",
              "  @keyframes spin {\n",
              "    0% {\n",
              "      border-color: transparent;\n",
              "      border-bottom-color: var(--fill-color);\n",
              "      border-left-color: var(--fill-color);\n",
              "    }\n",
              "    20% {\n",
              "      border-color: transparent;\n",
              "      border-left-color: var(--fill-color);\n",
              "      border-top-color: var(--fill-color);\n",
              "    }\n",
              "    30% {\n",
              "      border-color: transparent;\n",
              "      border-left-color: var(--fill-color);\n",
              "      border-top-color: var(--fill-color);\n",
              "      border-right-color: var(--fill-color);\n",
              "    }\n",
              "    40% {\n",
              "      border-color: transparent;\n",
              "      border-right-color: var(--fill-color);\n",
              "      border-top-color: var(--fill-color);\n",
              "    }\n",
              "    60% {\n",
              "      border-color: transparent;\n",
              "      border-right-color: var(--fill-color);\n",
              "    }\n",
              "    80% {\n",
              "      border-color: transparent;\n",
              "      border-right-color: var(--fill-color);\n",
              "      border-bottom-color: var(--fill-color);\n",
              "    }\n",
              "    90% {\n",
              "      border-color: transparent;\n",
              "      border-bottom-color: var(--fill-color);\n",
              "    }\n",
              "  }\n",
              "</style>\n",
              "\n",
              "  <script>\n",
              "    async function quickchart(key) {\n",
              "      const quickchartButtonEl =\n",
              "        document.querySelector('#' + key + ' button');\n",
              "      quickchartButtonEl.disabled = true;  // To prevent multiple clicks.\n",
              "      quickchartButtonEl.classList.add('colab-df-spinner');\n",
              "      try {\n",
              "        const charts = await google.colab.kernel.invokeFunction(\n",
              "            'suggestCharts', [key], {});\n",
              "      } catch (error) {\n",
              "        console.error('Error during call to suggestCharts:', error);\n",
              "      }\n",
              "      quickchartButtonEl.classList.remove('colab-df-spinner');\n",
              "      quickchartButtonEl.classList.add('colab-df-quickchart-complete');\n",
              "    }\n",
              "    (() => {\n",
              "      let quickchartButtonEl =\n",
              "        document.querySelector('#df-a6f312d0-4d04-4b75-96d1-f065fe203119 button');\n",
              "      quickchartButtonEl.style.display =\n",
              "        google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "    })();\n",
              "  </script>\n",
              "</div>\n",
              "    </div>\n",
              "  </div>\n"
            ]
          },
          "metadata": {},
          "execution_count": 19
        }
      ],
      "source": [
        "new_dfs.head()"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 20,
      "metadata": {
        "id": "DmnOPb0y-E-x"
      },
      "outputs": [],
      "source": [
        "new_dfs = new_dfs[['id', 'review']]\n",
        "neg_dfs = neg_dfs[['id', 'review']]\n"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 21,
      "metadata": {
        "id": "rAMJkLzm-E-x"
      },
      "outputs": [],
      "source": [
        "test_dfs = pd.merge(new_dfs, neg_dfs, how='outer')"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 22,
      "metadata": {
        "id": "KBpuJ60P-E-x",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 206
        },
        "outputId": "05ee6382-4f03-4a47-e352-0fadbc83e183"
      },
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "            id                                             review\n",
              "1506  27584915  This was literally the worst day that I’ve eve...\n",
              "518   32550619  the apartment looks new, and in very good loca...\n",
              "731   44747739  Great neighborhood close to Astoria park. Unfo...\n",
              "486   34487108                                                  .\n",
              "260   39837617  So nice and spacious place. nice host great ho..."
            ],
            "text/html": [
              "\n",
              "  <div id=\"df-a7cfd31e-1abf-4b00-ba8d-b1b7540fdbcc\" class=\"colab-df-container\">\n",
              "    <div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>id</th>\n",
              "      <th>review</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>1506</th>\n",
              "      <td>27584915</td>\n",
              "      <td>This was literally the worst day that I’ve eve...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>518</th>\n",
              "      <td>32550619</td>\n",
              "      <td>the apartment looks new, and in very good loca...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>731</th>\n",
              "      <td>44747739</td>\n",
              "      <td>Great neighborhood close to Astoria park. Unfo...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>486</th>\n",
              "      <td>34487108</td>\n",
              "      <td>.</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>260</th>\n",
              "      <td>39837617</td>\n",
              "      <td>So nice and spacious place. nice host great ho...</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>\n",
              "    <div class=\"colab-df-buttons\">\n",
              "\n",
              "  <div class=\"colab-df-container\">\n",
              "    <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-a7cfd31e-1abf-4b00-ba8d-b1b7540fdbcc')\"\n",
              "            title=\"Convert this dataframe to an interactive table.\"\n",
              "            style=\"display:none;\">\n",
              "\n",
              "  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\" viewBox=\"0 -960 960 960\">\n",
              "    <path d=\"M120-120v-720h720v720H120Zm60-500h600v-160H180v160Zm220 220h160v-160H400v160Zm0 220h160v-160H400v160ZM180-400h160v-160H180v160Zm440 0h160v-160H620v160ZM180-180h160v-160H180v160Zm440 0h160v-160H620v160Z\"/>\n",
              "  </svg>\n",
              "    </button>\n",
              "\n",
              "  <style>\n",
              "    .colab-df-container {\n",
              "      display:flex;\n",
              "      gap: 12px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert {\n",
              "      background-color: #E8F0FE;\n",
              "      border: none;\n",
              "      border-radius: 50%;\n",
              "      cursor: pointer;\n",
              "      display: none;\n",
              "      fill: #1967D2;\n",
              "      height: 32px;\n",
              "      padding: 0 0 0 0;\n",
              "      width: 32px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert:hover {\n",
              "      background-color: #E2EBFA;\n",
              "      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
              "      fill: #174EA6;\n",
              "    }\n",
              "\n",
              "    .colab-df-buttons div {\n",
              "      margin-bottom: 4px;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert {\n",
              "      background-color: #3B4455;\n",
              "      fill: #D2E3FC;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert:hover {\n",
              "      background-color: #434B5C;\n",
              "      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
              "      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
              "      fill: #FFFFFF;\n",
              "    }\n",
              "  </style>\n",
              "\n",
              "    <script>\n",
              "      const buttonEl =\n",
              "        document.querySelector('#df-a7cfd31e-1abf-4b00-ba8d-b1b7540fdbcc button.colab-df-convert');\n",
              "      buttonEl.style.display =\n",
              "        google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "\n",
              "      async function convertToInteractive(key) {\n",
              "        const element = document.querySelector('#df-a7cfd31e-1abf-4b00-ba8d-b1b7540fdbcc');\n",
              "        const dataTable =\n",
              "          await google.colab.kernel.invokeFunction('convertToInteractive',\n",
              "                                                    [key], {});\n",
              "        if (!dataTable) return;\n",
              "\n",
              "        const docLinkHtml = 'Like what you see? Visit the ' +\n",
              "          '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n",
              "          + ' to learn more about interactive tables.';\n",
              "        element.innerHTML = '';\n",
              "        dataTable['output_type'] = 'display_data';\n",
              "        await google.colab.output.renderOutput(dataTable, element);\n",
              "        const docLink = document.createElement('div');\n",
              "        docLink.innerHTML = docLinkHtml;\n",
              "        element.appendChild(docLink);\n",
              "      }\n",
              "    </script>\n",
              "  </div>\n",
              "\n",
              "\n",
              "<div id=\"df-c2fe32ec-68cd-4657-86cf-d0fd7ec09724\">\n",
              "  <button class=\"colab-df-quickchart\" onclick=\"quickchart('df-c2fe32ec-68cd-4657-86cf-d0fd7ec09724')\"\n",
              "            title=\"Suggest charts\"\n",
              "            style=\"display:none;\">\n",
              "\n",
              "<svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n",
              "     width=\"24px\">\n",
              "    <g>\n",
              "        <path d=\"M19 3H5c-1.1 0-2 .9-2 2v14c0 1.1.9 2 2 2h14c1.1 0 2-.9 2-2V5c0-1.1-.9-2-2-2zM9 17H7v-7h2v7zm4 0h-2V7h2v10zm4 0h-2v-4h2v4z\"/>\n",
              "    </g>\n",
              "</svg>\n",
              "  </button>\n",
              "\n",
              "<style>\n",
              "  .colab-df-quickchart {\n",
              "      --bg-color: #E8F0FE;\n",
              "      --fill-color: #1967D2;\n",
              "      --hover-bg-color: #E2EBFA;\n",
              "      --hover-fill-color: #174EA6;\n",
              "      --disabled-fill-color: #AAA;\n",
              "      --disabled-bg-color: #DDD;\n",
              "  }\n",
              "\n",
              "  [theme=dark] .colab-df-quickchart {\n",
              "      --bg-color: #3B4455;\n",
              "      --fill-color: #D2E3FC;\n",
              "      --hover-bg-color: #434B5C;\n",
              "      --hover-fill-color: #FFFFFF;\n",
              "      --disabled-bg-color: #3B4455;\n",
              "      --disabled-fill-color: #666;\n",
              "  }\n",
              "\n",
              "  .colab-df-quickchart {\n",
              "    background-color: var(--bg-color);\n",
              "    border: none;\n",
              "    border-radius: 50%;\n",
              "    cursor: pointer;\n",
              "    display: none;\n",
              "    fill: var(--fill-color);\n",
              "    height: 32px;\n",
              "    padding: 0;\n",
              "    width: 32px;\n",
              "  }\n",
              "\n",
              "  .colab-df-quickchart:hover {\n",
              "    background-color: var(--hover-bg-color);\n",
              "    box-shadow: 0 1px 2px rgba(60, 64, 67, 0.3), 0 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
              "    fill: var(--button-hover-fill-color);\n",
              "  }\n",
              "\n",
              "  .colab-df-quickchart-complete:disabled,\n",
              "  .colab-df-quickchart-complete:disabled:hover {\n",
              "    background-color: var(--disabled-bg-color);\n",
              "    fill: var(--disabled-fill-color);\n",
              "    box-shadow: none;\n",
              "  }\n",
              "\n",
              "  .colab-df-spinner {\n",
              "    border: 2px solid var(--fill-color);\n",
              "    border-color: transparent;\n",
              "    border-bottom-color: var(--fill-color);\n",
              "    animation:\n",
              "      spin 1s steps(1) infinite;\n",
              "  }\n",
              "\n",
              "  @keyframes spin {\n",
              "    0% {\n",
              "      border-color: transparent;\n",
              "      border-bottom-color: var(--fill-color);\n",
              "      border-left-color: var(--fill-color);\n",
              "    }\n",
              "    20% {\n",
              "      border-color: transparent;\n",
              "      border-left-color: var(--fill-color);\n",
              "      border-top-color: var(--fill-color);\n",
              "    }\n",
              "    30% {\n",
              "      border-color: transparent;\n",
              "      border-left-color: var(--fill-color);\n",
              "      border-top-color: var(--fill-color);\n",
              "      border-right-color: var(--fill-color);\n",
              "    }\n",
              "    40% {\n",
              "      border-color: transparent;\n",
              "      border-right-color: var(--fill-color);\n",
              "      border-top-color: var(--fill-color);\n",
              "    }\n",
              "    60% {\n",
              "      border-color: transparent;\n",
              "      border-right-color: var(--fill-color);\n",
              "    }\n",
              "    80% {\n",
              "      border-color: transparent;\n",
              "      border-right-color: var(--fill-color);\n",
              "      border-bottom-color: var(--fill-color);\n",
              "    }\n",
              "    90% {\n",
              "      border-color: transparent;\n",
              "      border-bottom-color: var(--fill-color);\n",
              "    }\n",
              "  }\n",
              "</style>\n",
              "\n",
              "  <script>\n",
              "    async function quickchart(key) {\n",
              "      const quickchartButtonEl =\n",
              "        document.querySelector('#' + key + ' button');\n",
              "      quickchartButtonEl.disabled = true;  // To prevent multiple clicks.\n",
              "      quickchartButtonEl.classList.add('colab-df-spinner');\n",
              "      try {\n",
              "        const charts = await google.colab.kernel.invokeFunction(\n",
              "            'suggestCharts', [key], {});\n",
              "      } catch (error) {\n",
              "        console.error('Error during call to suggestCharts:', error);\n",
              "      }\n",
              "      quickchartButtonEl.classList.remove('colab-df-spinner');\n",
              "      quickchartButtonEl.classList.add('colab-df-quickchart-complete');\n",
              "    }\n",
              "    (() => {\n",
              "      let quickchartButtonEl =\n",
              "        document.querySelector('#df-c2fe32ec-68cd-4657-86cf-d0fd7ec09724 button');\n",
              "      quickchartButtonEl.style.display =\n",
              "        google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "    })();\n",
              "  </script>\n",
              "</div>\n",
              "    </div>\n",
              "  </div>\n"
            ]
          },
          "metadata": {},
          "execution_count": 22
        }
      ],
      "source": [
        "test_dfs.sample(5)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 23,
      "metadata": {
        "id": "ilqb_gpk-E-x"
      },
      "outputs": [],
      "source": [
        "test_dfs.review = test_dfs.review.astype(str)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 24,
      "metadata": {
        "id": "kPSgBHa4-E-y",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "4ca98842-26c7-4f6d-c0b0-de76ea29d9a2"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "There are 1 GPU(s) available.\n",
            "Device name: Tesla T4\n"
          ]
        }
      ],
      "source": [
        "import torch\n",
        "\n",
        "if torch.cuda.is_available():\n",
        "    device = torch.device(\"cuda\")\n",
        "    print(f'There are {torch.cuda.device_count()} GPU(s) available.')\n",
        "    print('Device name:', torch.cuda.get_device_name(0))\n",
        "\n",
        "else:\n",
        "    print('No GPU available, using the CPU instead.')\n",
        "    device = torch.device(\"cpu\")"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 25,
      "metadata": {
        "id": "_3BvdRDY-E-y"
      },
      "outputs": [],
      "source": [
        "data = dfs.sample(100000)\n",
        "test_data = test_dfs"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 26,
      "metadata": {
        "id": "lOw1QOoZ-E-y"
      },
      "outputs": [],
      "source": [
        "from sklearn.model_selection import train_test_split\n",
        "\n",
        "X = data.review.values\n",
        "y = data.label.values\n",
        "\n",
        "X_train, X_val, y_train, y_val = train_test_split(X, y, test_size=0.1, random_state=2020)"
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "### TF-IDF + Naive Bayes Classifier\n",
        "\n"
      ],
      "metadata": {
        "id": "kpTxVyYWBYcX"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "import nltk\n",
        "# Uncomment to download \"stopwords\"\n",
        "nltk.download(\"stopwords\")\n",
        "from nltk.corpus import stopwords\n",
        "\n",
        "def text_preprocessing(s):\n",
        "    \"\"\"\n",
        "    - Lowercase the sentence\n",
        "    - Change \"'t\" to \"not\"\n",
        "    - Remove \"@name\"\n",
        "    - Isolate and remove punctuations except \"?\"\n",
        "    - Remove other special characters\n",
        "    - Remove stop words except \"not\" and \"can\"\n",
        "    - Remove trailing whitespace\n",
        "    \"\"\"\n",
        "    s = s.lower()\n",
        "    # Change 't to 'not'\n",
        "    s = re.sub(r\"\\'t\", \" not\", s)\n",
        "    # Remove @name\n",
        "    s = re.sub(r'(@.*?)[\\s]', ' ', s)\n",
        "    # Isolate and remove punctuations except '?'\n",
        "    s = re.sub(r'([\\'\\\"\\.\\(\\)\\!\\?\\\\\\/\\,])', r'  ', s)\n",
        "    s = re.sub(r'[^\\w\\s\\?]', ' ', s)\n",
        "    # Remove some special characters\n",
        "    s = re.sub(r'([\\;\\:\\|•«\\n])', ' ', s)\n",
        "    # Remove stopwords except 'not' and 'can'\n",
        "    s = \" \".join([word for word in s.split()\n",
        "                  if word not in stopwords.words('english')\n",
        "                  or word in ['not', 'can']])\n",
        "    # Remove trailing whitespace\n",
        "    s = re.sub(r'\\s+', ' ', s).strip()\n",
        "\n",
        "    return s"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "7QJobkJEBg3X",
        "outputId": "f8a63f25-2490-4b8a-d4f4-8b2d0bd18b6d"
      },
      "execution_count": 27,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "[nltk_data] Downloading package stopwords to /root/nltk_data...\n",
            "[nltk_data]   Unzipping corpora/stopwords.zip.\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "%%time\n",
        "from sklearn.feature_extraction.text import TfidfVectorizer\n",
        "\n",
        "# Preprocess text\n",
        "X_train_preprocessed = np.array([text_preprocessing(text) for text in X_train])\n",
        "X_val_preprocessed = np.array([text_preprocessing(text) for text in X_val])\n",
        "\n",
        "# Calculate TF-IDF\n",
        "tf_idf = TfidfVectorizer(smooth_idf=False)\n",
        "X_train_tfidf = tf_idf.fit_transform(X_train_preprocessed)\n",
        "X_val_tfidf = tf_idf.transform(X_val_preprocessed)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "_QPlBgJKBwX-",
        "outputId": "aa8a2e63-7af9-489b-991c-58a952c711e5"
      },
      "execution_count": 28,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "CPU times: user 6min 12s, sys: 53.8 s, total: 7min 6s\n",
            "Wall time: 7min 14s\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "from sklearn.model_selection import StratifiedKFold, cross_val_score\n",
        "\n",
        "def get_auc_CV(model):\n",
        "    \"\"\"\n",
        "    Return the average AUC score from cross-validation.\n",
        "    \"\"\"\n",
        "    # Set KFold to shuffle data before the split\n",
        "    kf = StratifiedKFold(5, shuffle=True, random_state=1)\n",
        "\n",
        "    # Get AUC scores\n",
        "    auc = cross_val_score(\n",
        "        model, X_train_tfidf, y_train, scoring=\"roc_auc\", cv=kf)\n",
        "\n",
        "    return auc.mean()"
      ],
      "metadata": {
        "id": "CHjVaXPoCLx5"
      },
      "execution_count": 32,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "from sklearn.naive_bayes import MultinomialNB\n",
        "\n",
        "alpha_values = np.arange(1, 10, 0.1)\n",
        "res = pd.Series([get_auc_CV(MultinomialNB(alpha=i)) for i in alpha_values],\n",
        "                index=alpha_values)\n",
        "best_alpha = np.round(res.idxmax(), 2)\n",
        "print('Best alpha: ', best_alpha)\n",
        "\n",
        "plt.plot(res)\n",
        "plt.title('AUC vs. Alpha')\n",
        "plt.xlabel('Alpha')\n",
        "plt.ylabel('AUC')\n",
        "plt.show()"
      ],
      "metadata": {
        "id": "8Zs4Fug6CvJg",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 490
        },
        "outputId": "4fc158de-34ce-4ce5-f239-659a18efa0be"
      },
      "execution_count": 33,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Best alpha:  1.0\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 640x480 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "from sklearn.metrics import accuracy_score, roc_curve, auc\n",
        "\n",
        "def evaluate_roc(probs, y_true):\n",
        "    \"\"\"\n",
        "    - Print AUC and accuracy on the test set\n",
        "    - Plot ROC\n",
        "    @params    probs (np.array): an array of predicted probabilities with shape (len(y_true), 2)\n",
        "    @params    y_true (np.array): an array of the true values with shape (len(y_true),)\n",
        "    \"\"\"\n",
        "    preds = probs[:, 1]\n",
        "    fpr, tpr, threshold = roc_curve(y_true, preds)\n",
        "    roc_auc = auc(fpr, tpr)\n",
        "    print(f'AUC: {roc_auc:.4f}')\n",
        "\n",
        "    # Get accuracy over the test set\n",
        "    y_pred = np.where(preds >= 0.5, 1, 0)\n",
        "    accuracy = accuracy_score(y_true, y_pred)\n",
        "    print(f'Accuracy: {accuracy*100:.2f}%')\n",
        "\n",
        "    # Plot ROC AUC\n",
        "    plt.title('Receiver Operating Characteristic')\n",
        "    plt.plot(fpr, tpr, 'b', label = 'AUC = %0.2f' % roc_auc)\n",
        "    plt.legend(loc = 'lower right')\n",
        "    plt.plot([0, 1], [0, 1],'r--')\n",
        "    plt.xlim([0, 1])\n",
        "    plt.ylim([0, 1])\n",
        "    plt.ylabel('True Positive Rate')\n",
        "    plt.xlabel('False Positive Rate')\n",
        "    plt.show()"
      ],
      "metadata": {
        "id": "EJ-4zNmlNPSe"
      },
      "execution_count": 34,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "# Compute predicted probabilities\n",
        "nb_model = MultinomialNB(alpha=1.0)\n",
        "nb_model.fit(X_train_tfidf, y_train)\n",
        "probs = nb_model.predict_proba(X_val_tfidf)\n",
        "\n",
        "# Evaluate the classifier\n",
        "evaluate_roc(probs, y_val)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 509
        },
        "id": "jCWCkvQrNUr2",
        "outputId": "83f1fb8f-2a6e-4e39-ff1f-d676c5ee3ff4"
      },
      "execution_count": 35,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "AUC: 0.5521\n",
            "Accuracy: 99.85%\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 640x480 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "### fine-tuning BERT"
      ],
      "metadata": {
        "id": "zOwhNucAShSI"
      }
    },
    {
      "cell_type": "code",
      "execution_count": 44,
      "metadata": {
        "id": "QQWeCu_--E-z"
      },
      "outputs": [],
      "source": [
        "def text_preprocessing(text):\n",
        "    \"\"\"\n",
        "    - Remove entity mentions (eg. '@')\n",
        "    - Correct errors\n",
        "    @param    text (str): a string to be processed.\n",
        "    @return   text (Str): the processed string.\n",
        "    \"\"\"\n",
        "    # Remove '@'\n",
        "    text = re.sub(r'(@.*?)[\\s]',' ', text)\n",
        "\n",
        "    # Replace '<\\br>' with ' '\n",
        "    text = re.sub(r'<\\br>;', ' ', text)\n",
        "\n",
        "    # Remove trailing whitespace\n",
        "    text = re.sub(r'\\s+', ' ', text).strip()\n",
        "\n",
        "    return text"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 45,
      "metadata": {
        "id": "7xuUgszU-E-z",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "1ba89433-a9f9-44ce-ba1f-7271b15f5d48"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Original:  Neil was a great host. Most of the time I travel, I bring my entire family and it’s very important to have a comfortable space that makes us feel at home. Neil’s place was just that. Very modern and comparable to our home. He was also very responsive to any additional need we had. The location is close to the highway and public transportation. Overall great place and very comparable to a luxury hotel. I can tell that Neil took the time to make sure every minor detail was added to the space.\n",
            "Processed:  Neil was a great host. Most of the time I travel, I bring my entire family and it’s very important to have a comfortable space that makes us feel at home. Neil’s place was just that. Very modern and comparable to our home. He was also very responsive to any additional need we had. The location is close to the highway and public transportation. Overall great place and very comparable to a luxury hotel. I can tell that Neil took the time to make sure every minor detail was added to the space.\n"
          ]
        }
      ],
      "source": [
        "# Print sentence 0\n",
        "print('Original: ', X[0])\n",
        "print('Processed: ', text_preprocessing(X[0]))"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 46,
      "metadata": {
        "id": "rXN4nbFZ-E-z"
      },
      "outputs": [],
      "source": [
        "from transformers import BertTokenizer\n",
        "\n",
        "# Load the BERT tokenizer\n",
        "tokenizer = BertTokenizer.from_pretrained('bert-base-uncased', do_lower_case=True)\n",
        "\n",
        "# Create a function to tokenize a set of texts\n",
        "def preprocessing_for_bert(data):\n",
        "    \"\"\"Perform required preprocessing steps for pretrained BERT.\n",
        "    @param    data (np.array): Array of texts to be processed.\n",
        "    @return   input_ids (torch.Tensor): Tensor of token ids to be fed to a model.\n",
        "    @return   attention_masks (torch.Tensor): Tensor of indices specifying which\n",
        "                  tokens should be attended to by the model.\n",
        "    \"\"\"\n",
        "    # Create empty lists to store outputs\n",
        "    input_ids = []\n",
        "    attention_masks = []\n",
        "\n",
        "    # For every sentence...\n",
        "    for sent in data:\n",
        "        # `encode_plus` will:\n",
        "        #    (1) Tokenize the sentence\n",
        "        #    (2) Add the `[CLS]` and `[SEP]` token to the start and end\n",
        "        #    (3) Truncate/Pad sentence to max length\n",
        "        #    (4) Map tokens to their IDs\n",
        "        #    (5) Create attention mask\n",
        "        #    (6) Return a dictionary of outputs\n",
        "        encoded_sent = tokenizer.encode_plus(\n",
        "            text=text_preprocessing(sent),  # Preprocess sentence\n",
        "            add_special_tokens=True,        # Add `[CLS]` and `[SEP]`\n",
        "            max_length=MAX_LEN,             # Max length to truncate/pad\n",
        "            pad_to_max_length=True,         # Pad sentence to max length\n",
        "            #return_tensors='pt',           # Return PyTorch tensor\n",
        "            return_attention_mask=True      # Return attention mask\n",
        "            )\n",
        "\n",
        "        # Add the outputs to the lists\n",
        "        input_ids.append(encoded_sent.get('input_ids'))\n",
        "        attention_masks.append(encoded_sent.get('attention_mask'))\n",
        "\n",
        "    # Convert lists to tensors\n",
        "    input_ids = torch.tensor(input_ids)\n",
        "    attention_masks = torch.tensor(attention_masks)\n",
        "\n",
        "    return input_ids, attention_masks"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 49,
      "metadata": {
        "id": "9j3FNXn--E-0",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "14684c84-ad20-492f-ab42-564b09746191"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Max length:  2343\n"
          ]
        }
      ],
      "source": [
        "# Concatenate train data and test data (test data is already a subset of train data)\n",
        "all_data = data.review.values\n",
        "\n",
        "# Encode our concatenated data\n",
        "encoded_data = [tokenizer.encode(sent, add_special_tokens=True) for sent in all_data]\n",
        "\n",
        "# Find the maximum length\n",
        "max_len = max([len(sent) for sent in encoded_data])\n",
        "print('Max length: ', max_len)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 50,
      "metadata": {
        "id": "4Bb_-_Lo-E-0",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "98e40eed-5e3e-4c32-fdb0-0290511d3c94"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Truncation was not explicitly activated but `max_length` is provided a specific value, please use `truncation=True` to explicitly truncate examples to max length. Defaulting to 'longest_first' truncation strategy. If you encode pairs of sequences (GLUE-style) with the tokenizer you can select this strategy more precisely by providing a specific strategy to `truncation`.\n",
            "/usr/local/lib/python3.10/dist-packages/transformers/tokenization_utils_base.py:2614: FutureWarning: The `pad_to_max_length` argument is deprecated and will be removed in a future version, use `padding=True` or `padding='longest'` to pad to the longest sequence in the batch, or use `padding='max_length'` to pad to a max length. In this case, you can give a specific length with `max_length` (e.g. `max_length=45`) or leave max_length to None to pad to the maximal input size of the model (e.g. 512 for Bert).\n",
            "  warnings.warn(\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Original:  Neil was a great host. Most of the time I travel, I bring my entire family and it’s very important to have a comfortable space that makes us feel at home. Neil’s place was just that. Very modern and comparable to our home. He was also very responsive to any additional need we had. The location is close to the highway and public transportation. Overall great place and very comparable to a luxury hotel. I can tell that Neil took the time to make sure every minor detail was added to the space.\n",
            "Token IDs:  [101, 6606, 2001, 1037, 2307, 3677, 1012, 2087, 1997, 1996, 2051, 1045, 3604, 1010, 1045, 3288, 2026, 2972, 2155, 1998, 2009, 1521, 1055, 2200, 2590, 2000, 2031, 1037, 6625, 2686, 2008, 3084, 2149, 2514, 2012, 2188, 1012, 6606, 1521, 1055, 2173, 2001, 2074, 2008, 1012, 2200, 2715, 1998, 12435, 2000, 2256, 2188, 1012, 2002, 2001, 2036, 2200, 26651, 2000, 2151, 3176, 2342, 2057, 102]\n",
            "Tokenizing data...\n"
          ]
        }
      ],
      "source": [
        "# Specify `MAX_LEN`\n",
        "MAX_LEN = 64\n",
        "\n",
        "# Print sentence 0 and its encoded token ids\n",
        "token_ids = list(preprocessing_for_bert([X[0]])[0].squeeze().numpy())\n",
        "print('Original: ', X[0])\n",
        "print('Token IDs: ', token_ids)\n",
        "\n",
        "# Run function `preprocessing_for_bert` on the train set and the validation set\n",
        "print('Tokenizing data...')\n",
        "train_inputs, train_masks = preprocessing_for_bert(X_train)\n",
        "val_inputs, val_masks = preprocessing_for_bert(X_val)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 51,
      "metadata": {
        "id": "svydY_ta-E-0"
      },
      "outputs": [],
      "source": [
        "from torch.utils.data import TensorDataset, DataLoader, RandomSampler, SequentialSampler\n",
        "\n",
        "# Convert other data types to torch.Tensor\n",
        "train_labels = torch.tensor(y_train)\n",
        "val_labels = torch.tensor(y_val)\n",
        "\n",
        "# For fine-tuning BERT, the authors recommend a batch size of 16 or 32.\n",
        "batch_size = 32\n",
        "\n",
        "# Create the DataLoader for our training set\n",
        "train_data = TensorDataset(train_inputs, train_masks, train_labels)\n",
        "train_sampler = RandomSampler(train_data)\n",
        "train_dataloader = DataLoader(train_data, sampler=train_sampler, batch_size=batch_size)\n",
        "\n",
        "# Create the DataLoader for our validation set\n",
        "val_data = TensorDataset(val_inputs, val_masks, val_labels)\n",
        "val_sampler = SequentialSampler(val_data)\n",
        "val_dataloader = DataLoader(val_data, sampler=val_sampler, batch_size=batch_size)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 52,
      "metadata": {
        "id": "fcloJVvo-E-0",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "22ee54b6-aaa1-4acf-ba8f-466f1cb307b9"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "CPU times: user 78 ms, sys: 4.69 ms, total: 82.7 ms\n",
            "Wall time: 114 ms\n"
          ]
        }
      ],
      "source": [
        "%%time\n",
        "import torch\n",
        "import torch.nn as nn\n",
        "from transformers import BertModel\n",
        "\n",
        "# Create the BertClassfier class\n",
        "class BertClassifier(nn.Module):\n",
        "    \"\"\"Bert Model for Classification Tasks.\n",
        "    \"\"\"\n",
        "    def __init__(self, freeze_bert=False):\n",
        "        \"\"\"\n",
        "        @param    bert: a BertModel object\n",
        "        @param    classifier: a torch.nn.Module classifier\n",
        "        @param    freeze_bert (bool): Set `False` to fine-tune the BERT model\n",
        "        \"\"\"\n",
        "        super(BertClassifier, self).__init__()\n",
        "        # Specify hidden size of BERT, hidden size of our classifier, and number of labels\n",
        "        D_in, H, D_out = 768, 50, 2\n",
        "\n",
        "        # Instantiate BERT model\n",
        "        self.bert = BertModel.from_pretrained('bert-base-uncased')\n",
        "\n",
        "        # Instantiate an one-layer feed-forward classifier\n",
        "        self.classifier = nn.Sequential(\n",
        "            nn.Linear(D_in, H),\n",
        "            nn.ReLU(),\n",
        "            #nn.Dropout(0.5),\n",
        "            nn.Linear(H, D_out)\n",
        "        )\n",
        "\n",
        "        # Freeze the BERT model\n",
        "        if freeze_bert:\n",
        "            for param in self.bert.parameters():\n",
        "                param.requires_grad = False\n",
        "\n",
        "    def forward(self, input_ids, attention_mask):\n",
        "        \"\"\"\n",
        "        Feed input to BERT and the classifier to compute logits.\n",
        "        @param    input_ids (torch.Tensor): an input tensor with shape (batch_size,\n",
        "                      max_length)\n",
        "        @param    attention_mask (torch.Tensor): a tensor that hold attention mask\n",
        "                      information with shape (batch_size, max_length)\n",
        "        @return   logits (torch.Tensor): an output tensor with shape (batch_size,\n",
        "                      num_labels)\n",
        "        \"\"\"\n",
        "        # Feed input to BERT\n",
        "        outputs = self.bert(input_ids=input_ids,\n",
        "                            attention_mask=attention_mask)\n",
        "\n",
        "        # Extract the last hidden state of the token `[CLS]` for classification task\n",
        "        last_hidden_state_cls = outputs[0][:, 0, :]\n",
        "\n",
        "        # Feed input to classifier to compute logits\n",
        "        logits = self.classifier(last_hidden_state_cls)\n",
        "\n",
        "        return logits"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 57,
      "metadata": {
        "id": "S7mP3nBD-E-1"
      },
      "outputs": [],
      "source": [
        "from transformers import AdamW, get_linear_schedule_with_warmup\n",
        "\n",
        "def initialize_model(epochs=4):\n",
        "    \"\"\"Initialize the Bert Classifier, the optimizer and the learning rate scheduler.\n",
        "    \"\"\"\n",
        "    # Instantiate Bert Classifier\n",
        "    bert_classifier = BertClassifier(freeze_bert=False)\n",
        "\n",
        "    # Tell PyTorch to run the model on GPU\n",
        "    bert_classifier.to(device)\n",
        "\n",
        "    # Create the optimizer\n",
        "    optimizer = torch.optim.AdamW(bert_classifier.parameters(),\n",
        "                      lr=5e-5,    # Default learning rate\n",
        "                      eps=1e-8    # Default epsilon value\n",
        "                      )\n",
        "\n",
        "    # Total number of training steps\n",
        "    total_steps = len(train_dataloader) * epochs\n",
        "\n",
        "    # Set up the learning rate scheduler\n",
        "    scheduler = get_linear_schedule_with_warmup(optimizer,\n",
        "                                                num_warmup_steps=0, # Default value\n",
        "                                                num_training_steps=total_steps)\n",
        "    return bert_classifier, optimizer, scheduler"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 58,
      "metadata": {
        "id": "Asab23gm-E-1"
      },
      "outputs": [],
      "source": [
        "import random\n",
        "import time\n",
        "\n",
        "# Specify loss function\n",
        "loss_fn = nn.CrossEntropyLoss()\n",
        "\n",
        "def set_seed(seed_value=42):\n",
        "    \"\"\"Set seed for reproducibility.\n",
        "    \"\"\"\n",
        "    random.seed(seed_value)\n",
        "    np.random.seed(seed_value)\n",
        "    torch.manual_seed(seed_value)\n",
        "    torch.cuda.manual_seed_all(seed_value)\n",
        "\n",
        "def train(model, train_dataloader, val_dataloader=None, epochs=4, evaluation=False):\n",
        "    \"\"\"Train the BertClassifier model.\n",
        "    \"\"\"\n",
        "    # Start training loop\n",
        "    print(\"Start training...\\n\")\n",
        "    for epoch_i in range(epochs):\n",
        "        # =======================================\n",
        "        #               Training\n",
        "        # =======================================\n",
        "        # Print the header of the result table\n",
        "        print(f\"{'Epoch':^7} | {'Batch':^7} | {'Train Loss':^12} | {'Val Loss':^10} | {'Val Acc':^9} | {'Elapsed':^9}\")\n",
        "        print(\"-\"*70)\n",
        "\n",
        "        # Measure the elapsed time of each epoch\n",
        "        t0_epoch, t0_batch = time.time(), time.time()\n",
        "\n",
        "        # Reset tracking variables at the beginning of each epoch\n",
        "        total_loss, batch_loss, batch_counts = 0, 0, 0\n",
        "\n",
        "        # Put the model into the training mode\n",
        "        model.train()\n",
        "\n",
        "        # For each batch of training data...\n",
        "        for step, batch in enumerate(train_dataloader):\n",
        "            batch_counts +=1\n",
        "            # Load batch to GPU\n",
        "            b_input_ids, b_attn_mask, b_labels = tuple(t.to(device) for t in batch)\n",
        "\n",
        "            # Zero out any previously calculated gradients\n",
        "            model.zero_grad()\n",
        "\n",
        "            # Perform a forward pass. This will return logits.\n",
        "            logits = model(b_input_ids, b_attn_mask)\n",
        "\n",
        "            # Compute loss and accumulate the loss values\n",
        "            loss = loss_fn(logits, b_labels)\n",
        "            batch_loss += loss.item()\n",
        "            total_loss += loss.item()\n",
        "\n",
        "            # Perform a backward pass to calculate gradients\n",
        "            loss.backward()\n",
        "\n",
        "            # Clip the norm of the gradients to 1.0 to prevent \"exploding gradients\"\n",
        "            torch.nn.utils.clip_grad_norm_(model.parameters(), 1.0)\n",
        "\n",
        "            # Update parameters and the learning rate\n",
        "            optimizer.step()\n",
        "            scheduler.step()\n",
        "\n",
        "            # Print the loss values and time elapsed for every 20 batches\n",
        "            if (step % 20 == 0 and step != 0) or (step == len(train_dataloader) - 1):\n",
        "                # Calculate time elapsed for 20 batches\n",
        "                time_elapsed = time.time() - t0_batch\n",
        "\n",
        "                # Print training results\n",
        "                print(f\"{epoch_i + 1:^7} | {step:^7} | {batch_loss / batch_counts:^12.6f} | {'-':^10} | {'-':^9} | {time_elapsed:^9.2f}\")\n",
        "\n",
        "                # Reset batch tracking variables\n",
        "                batch_loss, batch_counts = 0, 0\n",
        "                t0_batch = time.time()\n",
        "\n",
        "        # Calculate the average loss over the entire training data\n",
        "        avg_train_loss = total_loss / len(train_dataloader)\n",
        "\n",
        "        print(\"-\"*70)\n",
        "        # =======================================\n",
        "        #               Evaluation\n",
        "        # =======================================\n",
        "        if evaluation == True:\n",
        "            # After the completion of each training epoch, measure the model's performance\n",
        "            # on our validation set.\n",
        "            val_loss, val_accuracy = evaluate(model, val_dataloader)\n",
        "\n",
        "            # Print performance over the entire training data\n",
        "            time_elapsed = time.time() - t0_epoch\n",
        "\n",
        "            print(f\"{epoch_i + 1:^7} | {'-':^7} | {avg_train_loss:^12.6f} | {val_loss:^10.6f} | {val_accuracy:^9.2f} | {time_elapsed:^9.2f}\")\n",
        "            print(\"-\"*70)\n",
        "        print(\"\\n\")\n",
        "\n",
        "    print(\"Training complete!\")\n",
        "\n",
        "\n",
        "def evaluate(model, val_dataloader):\n",
        "    \"\"\"After the completion of each training epoch, measure the model's performance\n",
        "    on our validation set.\n",
        "    \"\"\"\n",
        "    # Put the model into the evaluation mode. The dropout layers are disabled during\n",
        "    # the test time.\n",
        "    model.eval()\n",
        "\n",
        "    # Tracking variables\n",
        "    val_accuracy = []\n",
        "    val_loss = []\n",
        "\n",
        "    # For each batch in our validation set...\n",
        "    for batch in val_dataloader:\n",
        "        # Load batch to GPU\n",
        "        b_input_ids, b_attn_mask, b_labels = tuple(t.to(device) for t in batch)\n",
        "\n",
        "        # Compute logits\n",
        "        with torch.no_grad():\n",
        "            logits = model(b_input_ids, b_attn_mask)\n",
        "\n",
        "        # Compute loss\n",
        "        loss = loss_fn(logits, b_labels)\n",
        "        val_loss.append(loss.item())\n",
        "\n",
        "        # Get the predictions\n",
        "        preds = torch.argmax(logits, dim=1).flatten()\n",
        "\n",
        "        # Calculate the accuracy rate\n",
        "        accuracy = (preds == b_labels).cpu().numpy().mean() * 100\n",
        "        val_accuracy.append(accuracy)\n",
        "\n",
        "    # Compute the average accuracy and loss over the validation set.\n",
        "    val_loss = np.mean(val_loss)\n",
        "    val_accuracy = np.mean(val_accuracy)\n",
        "\n",
        "    return val_loss, val_accuracy"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 59,
      "metadata": {
        "id": "QafG5CJa-E-1",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "5089d37b-ca2d-410f-9b61-4effd64df73d"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Start training...\n",
            "\n",
            " Epoch  |  Batch  |  Train Loss  |  Val Loss  |  Val Acc  |  Elapsed \n",
            "----------------------------------------------------------------------\n",
            "   1    |   20    |   0.101505   |     -      |     -     |   6.59   \n",
            "   1    |   40    |   0.012489   |     -      |     -     |   6.16   \n",
            "   1    |   60    |   0.021365   |     -      |     -     |   6.22   \n",
            "   1    |   80    |   0.002581   |     -      |     -     |   6.27   \n",
            "   1    |   100   |   0.011822   |     -      |     -     |   6.34   \n",
            "   1    |   120   |   0.011839   |     -      |     -     |   6.35   \n",
            "   1    |   140   |   0.001902   |     -      |     -     |   6.41   \n",
            "   1    |   160   |   0.001471   |     -      |     -     |   6.49   \n",
            "   1    |   180   |   0.001196   |     -      |     -     |   6.48   \n",
            "   1    |   200   |   0.011655   |     -      |     -     |   6.41   \n",
            "   1    |   220   |   0.001243   |     -      |     -     |   6.45   \n",
            "   1    |   240   |   0.012051   |     -      |     -     |   6.48   \n",
            "   1    |   260   |   0.011982   |     -      |     -     |   6.53   \n",
            "   1    |   280   |   0.001047   |     -      |     -     |   6.50   \n",
            "   1    |   300   |   0.000834   |     -      |     -     |   6.57   \n",
            "   1    |   320   |   0.011804   |     -      |     -     |   6.59   \n",
            "   1    |   340   |   0.000887   |     -      |     -     |   6.68   \n",
            "   1    |   360   |   0.012140   |     -      |     -     |   6.63   \n",
            "   1    |   380   |   0.011969   |     -      |     -     |   6.65   \n",
            "   1    |   400   |   0.012182   |     -      |     -     |   6.72   \n",
            "   1    |   420   |   0.011568   |     -      |     -     |   6.73   \n",
            "   1    |   440   |   0.001135   |     -      |     -     |   6.78   \n",
            "   1    |   460   |   0.033443   |     -      |     -     |   6.78   \n",
            "   1    |   480   |   0.002165   |     -      |     -     |   6.83   \n",
            "   1    |   500   |   0.001006   |     -      |     -     |   6.84   \n",
            "   1    |   520   |   0.012032   |     -      |     -     |   6.85   \n",
            "   1    |   540   |   0.012068   |     -      |     -     |   6.88   \n",
            "   1    |   560   |   0.012361   |     -      |     -     |   6.82   \n",
            "   1    |   580   |   0.001127   |     -      |     -     |   6.83   \n",
            "   1    |   600   |   0.012109   |     -      |     -     |   6.97   \n",
            "   1    |   620   |   0.034070   |     -      |     -     |   6.88   \n",
            "   1    |   640   |   0.001884   |     -      |     -     |   6.87   \n",
            "   1    |   660   |   0.001139   |     -      |     -     |   6.89   \n",
            "   1    |   680   |   0.000887   |     -      |     -     |   6.98   \n",
            "   1    |   700   |   0.000750   |     -      |     -     |   6.98   \n",
            "   1    |   720   |   0.000658   |     -      |     -     |   6.86   \n",
            "   1    |   740   |   0.012174   |     -      |     -     |   6.91   \n",
            "   1    |   760   |   0.012121   |     -      |     -     |   6.90   \n",
            "   1    |   780   |   0.000876   |     -      |     -     |   6.92   \n",
            "   1    |   800   |   0.000761   |     -      |     -     |   6.93   \n",
            "   1    |   820   |   0.012107   |     -      |     -     |   6.92   \n",
            "   1    |   840   |   0.012281   |     -      |     -     |   6.94   \n",
            "   1    |   860   |   0.000934   |     -      |     -     |   6.95   \n",
            "   1    |   880   |   0.012104   |     -      |     -     |   6.96   \n",
            "   1    |   900   |   0.000821   |     -      |     -     |   6.99   \n",
            "   1    |   920   |   0.012168   |     -      |     -     |   6.97   \n",
            "   1    |   940   |   0.012136   |     -      |     -     |   7.03   \n",
            "   1    |   960   |   0.000918   |     -      |     -     |   7.05   \n",
            "   1    |   980   |   0.012161   |     -      |     -     |   7.06   \n",
            "   1    |  1000   |   0.022918   |     -      |     -     |   7.08   \n",
            "   1    |  1020   |   0.001395   |     -      |     -     |   7.11   \n",
            "   1    |  1040   |   0.000933   |     -      |     -     |   7.09   \n",
            "   1    |  1060   |   0.000769   |     -      |     -     |   7.07   \n",
            "   1    |  1080   |   0.012141   |     -      |     -     |   7.06   \n",
            "   1    |  1100   |   0.000795   |     -      |     -     |   7.06   \n",
            "   1    |  1120   |   0.000673   |     -      |     -     |   7.08   \n",
            "   1    |  1140   |   0.012310   |     -      |     -     |   7.02   \n",
            "   1    |  1160   |   0.022663   |     -      |     -     |   7.04   \n",
            "   1    |  1180   |   0.038866   |     -      |     -     |   7.05   \n",
            "   1    |  1200   |   0.013532   |     -      |     -     |   7.06   \n",
            "   1    |  1220   |   0.001448   |     -      |     -     |   7.04   \n",
            "   1    |  1240   |   0.012077   |     -      |     -     |   7.09   \n",
            "   1    |  1260   |   0.022256   |     -      |     -     |   7.06   \n",
            "   1    |  1280   |   0.001739   |     -      |     -     |   7.09   \n",
            "   1    |  1300   |   0.011941   |     -      |     -     |   7.07   \n",
            "   1    |  1320   |   0.011929   |     -      |     -     |   7.08   \n",
            "   1    |  1340   |   0.011853   |     -      |     -     |   7.08   \n",
            "   1    |  1360   |   0.011853   |     -      |     -     |   7.15   \n",
            "   1    |  1380   |   0.012417   |     -      |     -     |   7.08   \n",
            "   1    |  1400   |   0.001265   |     -      |     -     |   7.10   \n",
            "   1    |  1420   |   0.011854   |     -      |     -     |   7.11   \n",
            "   1    |  1440   |   0.032796   |     -      |     -     |   7.08   \n",
            "   1    |  1460   |   0.002345   |     -      |     -     |   7.11   \n",
            "   1    |  1480   |   0.001304   |     -      |     -     |   7.09   \n",
            "   1    |  1500   |   0.011864   |     -      |     -     |   7.11   \n",
            "   1    |  1520   |   0.011815   |     -      |     -     |   7.10   \n",
            "   1    |  1540   |   0.011847   |     -      |     -     |   7.08   \n",
            "   1    |  1560   |   0.001238   |     -      |     -     |   7.08   \n",
            "   1    |  1580   |   0.000972   |     -      |     -     |   7.09   \n",
            "   1    |  1600   |   0.011981   |     -      |     -     |   7.09   \n",
            "   1    |  1620   |   0.012003   |     -      |     -     |   7.07   \n",
            "   1    |  1640   |   0.033557   |     -      |     -     |   7.07   \n",
            "   1    |  1660   |   0.030901   |     -      |     -     |   7.07   \n",
            "   1    |  1680   |   0.002869   |     -      |     -     |   7.08   \n",
            "   1    |  1700   |   0.001527   |     -      |     -     |   7.07   \n",
            "   1    |  1720   |   0.022144   |     -      |     -     |   7.06   \n",
            "   1    |  1740   |   0.001435   |     -      |     -     |   7.10   \n",
            "   1    |  1760   |   0.022405   |     -      |     -     |   7.07   \n",
            "   1    |  1780   |   0.011862   |     -      |     -     |   7.10   \n",
            "   1    |  1800   |   0.001511   |     -      |     -     |   7.12   \n",
            "   1    |  1820   |   0.011822   |     -      |     -     |   7.21   \n",
            "   1    |  1840   |   0.011782   |     -      |     -     |   7.26   \n",
            "   1    |  1860   |   0.022276   |     -      |     -     |   7.24   \n",
            "   1    |  1880   |   0.011745   |     -      |     -     |   7.24   \n",
            "   1    |  1900   |   0.001552   |     -      |     -     |   7.20   \n",
            "   1    |  1920   |   0.001243   |     -      |     -     |   7.16   \n",
            "   1    |  1940   |   0.001029   |     -      |     -     |   7.14   \n",
            "   1    |  1960   |   0.011992   |     -      |     -     |   7.11   \n",
            "   1    |  1980   |   0.011795   |     -      |     -     |   7.08   \n",
            "   1    |  2000   |   0.001048   |     -      |     -     |   7.08   \n",
            "   1    |  2020   |   0.000908   |     -      |     -     |   7.12   \n",
            "   1    |  2040   |   0.000804   |     -      |     -     |   7.10   \n",
            "   1    |  2060   |   0.023264   |     -      |     -     |   7.12   \n",
            "   1    |  2080   |   0.022713   |     -      |     -     |   7.14   \n",
            "   1    |  2100   |   0.012073   |     -      |     -     |   7.14   \n",
            "   1    |  2120   |   0.011673   |     -      |     -     |   7.14   \n",
            "   1    |  2140   |   0.011610   |     -      |     -     |   7.17   \n",
            "   1    |  2160   |   0.011794   |     -      |     -     |   7.17   \n",
            "   1    |  2180   |   0.001299   |     -      |     -     |   7.11   \n",
            "   1    |  2200   |   0.022488   |     -      |     -     |   7.16   \n",
            "   1    |  2220   |   0.011804   |     -      |     -     |   7.13   \n",
            "   1    |  2240   |   0.011575   |     -      |     -     |   7.14   \n",
            "   1    |  2260   |   0.001397   |     -      |     -     |   7.15   \n",
            "   1    |  2280   |   0.001131   |     -      |     -     |   7.11   \n",
            "   1    |  2300   |   0.000972   |     -      |     -     |   7.11   \n",
            "   1    |  2320   |   0.011907   |     -      |     -     |   7.13   \n",
            "   1    |  2340   |   0.011811   |     -      |     -     |   7.14   \n",
            "   1    |  2360   |   0.001006   |     -      |     -     |   7.10   \n",
            "   1    |  2380   |   0.011984   |     -      |     -     |   7.09   \n",
            "   1    |  2400   |   0.000933   |     -      |     -     |   7.11   \n",
            "   1    |  2420   |   0.044462   |     -      |     -     |   7.12   \n",
            "   1    |  2440   |   0.001699   |     -      |     -     |   7.08   \n",
            "   1    |  2460   |   0.011643   |     -      |     -     |   7.11   \n",
            "   1    |  2480   |   0.022538   |     -      |     -     |   7.10   \n",
            "   1    |  2500   |   0.001481   |     -      |     -     |   7.09   \n",
            "   1    |  2520   |   0.011776   |     -      |     -     |   7.09   \n",
            "   1    |  2540   |   0.022225   |     -      |     -     |   7.09   \n",
            "   1    |  2560   |   0.011749   |     -      |     -     |   7.10   \n",
            "   1    |  2580   |   0.011859   |     -      |     -     |   7.10   \n",
            "   1    |  2600   |   0.011844   |     -      |     -     |   7.11   \n",
            "   1    |  2620   |   0.011582   |     -      |     -     |   7.10   \n",
            "   1    |  2640   |   0.011751   |     -      |     -     |   7.18   \n",
            "   1    |  2660   |   0.001562   |     -      |     -     |   7.20   \n",
            "   1    |  2680   |   0.011697   |     -      |     -     |   7.20   \n",
            "   1    |  2700   |   0.011696   |     -      |     -     |   7.23   \n",
            "   1    |  2720   |   0.021923   |     -      |     -     |   7.21   \n",
            "   1    |  2740   |   0.011835   |     -      |     -     |   7.17   \n",
            "   1    |  2760   |   0.001543   |     -      |     -     |   7.14   \n",
            "   1    |  2780   |   0.011787   |     -      |     -     |   7.13   \n",
            "   1    |  2800   |   0.032189   |     -      |     -     |   7.12   \n",
            "   1    |  2812   |   0.018158   |     -      |     -     |   4.24   \n",
            "----------------------------------------------------------------------\n",
            "   1    |    -    |   0.010822   |  0.009971  |   99.87   |  1019.37 \n",
            "----------------------------------------------------------------------\n",
            "\n",
            "\n",
            " Epoch  |  Batch  |  Train Loss  |  Val Loss  |  Val Acc  |  Elapsed \n",
            "----------------------------------------------------------------------\n",
            "   2    |   20    |   0.002110   |     -      |     -     |   7.25   \n",
            "   2    |   40    |   0.001633   |     -      |     -     |   7.12   \n",
            "   2    |   60    |   0.011779   |     -      |     -     |   7.13   \n",
            "   2    |   80    |   0.001195   |     -      |     -     |   7.09   \n",
            "   2    |   100   |   0.011692   |     -      |     -     |   7.12   \n",
            "   2    |   120   |   0.011868   |     -      |     -     |   7.16   \n",
            "   2    |   140   |   0.011809   |     -      |     -     |   7.11   \n",
            "   2    |   160   |   0.001239   |     -      |     -     |   7.14   \n",
            "   2    |   180   |   0.011864   |     -      |     -     |   7.14   \n",
            "   2    |   200   |   0.022427   |     -      |     -     |   7.15   \n",
            "   2    |   220   |   0.001390   |     -      |     -     |   7.11   \n",
            "   2    |   240   |   0.011662   |     -      |     -     |   7.16   \n",
            "   2    |   260   |   0.022333   |     -      |     -     |   7.11   \n",
            "   2    |   280   |   0.001398   |     -      |     -     |   7.14   \n",
            "   2    |   300   |   0.001293   |     -      |     -     |   7.11   \n",
            "   2    |   320   |   0.011720   |     -      |     -     |   7.22   \n",
            "   2    |   340   |   0.011792   |     -      |     -     |   7.15   \n",
            "   2    |   360   |   0.001200   |     -      |     -     |   7.21   \n",
            "   2    |   380   |   0.011786   |     -      |     -     |   7.13   \n",
            "   2    |   400   |   0.001121   |     -      |     -     |   7.14   \n",
            "   2    |   420   |   0.032944   |     -      |     -     |   7.12   \n",
            "   2    |   440   |   0.001400   |     -      |     -     |   7.12   \n",
            "   2    |   460   |   0.001260   |     -      |     -     |   7.18   \n",
            "   2    |   480   |   0.011774   |     -      |     -     |   7.11   \n",
            "   2    |   500   |   0.011896   |     -      |     -     |   7.11   \n",
            "   2    |   520   |   0.001186   |     -      |     -     |   7.10   \n",
            "   2    |   540   |   0.022374   |     -      |     -     |   7.16   \n",
            "   2    |   560   |   0.011687   |     -      |     -     |   7.14   \n",
            "   2    |   580   |   0.001316   |     -      |     -     |   7.17   \n",
            "   2    |   600   |   0.001170   |     -      |     -     |   7.19   \n",
            "   2    |   620   |   0.011776   |     -      |     -     |   7.20   \n",
            "   2    |   640   |   0.022517   |     -      |     -     |   7.23   \n",
            "   2    |   660   |   0.011794   |     -      |     -     |   7.16   \n",
            "   2    |   680   |   0.021828   |     -      |     -     |   7.15   \n",
            "   2    |   700   |   0.001630   |     -      |     -     |   7.11   \n",
            "   2    |   720   |   0.011655   |     -      |     -     |   7.09   \n",
            "   2    |   740   |   0.001384   |     -      |     -     |   7.12   \n",
            "   2    |   760   |   0.001193   |     -      |     -     |   7.07   \n",
            "   2    |   780   |   0.011740   |     -      |     -     |   7.12   \n",
            "   2    |   800   |   0.001055   |     -      |     -     |   7.09   \n",
            "   2    |   820   |   0.001013   |     -      |     -     |   7.10   \n",
            "   2    |   840   |   0.000933   |     -      |     -     |   7.11   \n",
            "   2    |   860   |   0.011918   |     -      |     -     |   7.10   \n",
            "   2    |   880   |   0.000918   |     -      |     -     |   7.11   \n",
            "   2    |   900   |   0.022844   |     -      |     -     |   7.13   \n",
            "   2    |   920   |   0.000984   |     -      |     -     |   7.12   \n",
            "   2    |   940   |   0.011871   |     -      |     -     |   7.13   \n",
            "   2    |   960   |   0.001006   |     -      |     -     |   7.13   \n",
            "   2    |   980   |   0.000930   |     -      |     -     |   7.11   \n",
            "   2    |  1000   |   0.000860   |     -      |     -     |   7.17   \n",
            "   2    |  1020   |   0.000801   |     -      |     -     |   7.12   \n",
            "   2    |  1040   |   0.012040   |     -      |     -     |   7.14   \n",
            "   2    |  1060   |   0.034194   |     -      |     -     |   7.12   \n",
            "   2    |  1080   |   0.021925   |     -      |     -     |   7.11   \n",
            "   2    |  1100   |   0.022134   |     -      |     -     |   7.13   \n",
            "   2    |  1120   |   0.001459   |     -      |     -     |   7.09   \n",
            "   2    |  1140   |   0.011716   |     -      |     -     |   7.12   \n",
            "   2    |  1160   |   0.032299   |     -      |     -     |   7.10   \n",
            "   2    |  1180   |   0.011838   |     -      |     -     |   7.11   \n",
            "   2    |  1200   |   0.001756   |     -      |     -     |   7.10   \n",
            "   2    |  1220   |   0.001420   |     -      |     -     |   7.10   \n",
            "   2    |  1240   |   0.001211   |     -      |     -     |   7.12   \n",
            "   2    |  1260   |   0.011719   |     -      |     -     |   7.10   \n",
            "   2    |  1280   |   0.011705   |     -      |     -     |   7.09   \n",
            "   2    |  1300   |   0.011507   |     -      |     -     |   7.08   \n",
            "   2    |  1320   |   0.011941   |     -      |     -     |   7.08   \n",
            "   2    |  1340   |   0.011414   |     -      |     -     |   7.10   \n",
            "   2    |  1360   |   0.001258   |     -      |     -     |   7.11   \n",
            "   2    |  1380   |   0.011890   |     -      |     -     |   7.12   \n",
            "   2    |  1400   |   0.020812   |     -      |     -     |   7.15   \n",
            "   2    |  1420   |   0.001412   |     -      |     -     |   7.19   \n",
            "   2    |  1440   |   0.019311   |     -      |     -     |   7.21   \n",
            "   2    |  1460   |   0.012006   |     -      |     -     |   7.22   \n",
            "   2    |  1480   |   0.021738   |     -      |     -     |   7.19   \n",
            "   2    |  1500   |   0.032022   |     -      |     -     |   7.16   \n",
            "   2    |  1520   |   0.002301   |     -      |     -     |   7.15   \n",
            "   2    |  1540   |   0.011804   |     -      |     -     |   7.13   \n",
            "   2    |  1560   |   0.001366   |     -      |     -     |   7.12   \n",
            "   2    |  1580   |   0.021656   |     -      |     -     |   7.11   \n",
            "   2    |  1600   |   0.012027   |     -      |     -     |   7.10   \n",
            "   2    |  1620   |   0.001512   |     -      |     -     |   7.08   \n",
            "   2    |  1640   |   0.001293   |     -      |     -     |   7.12   \n",
            "   2    |  1660   |   0.011791   |     -      |     -     |   7.13   \n",
            "   2    |  1680   |   0.001185   |     -      |     -     |   7.14   \n",
            "   2    |  1700   |   0.022497   |     -      |     -     |   7.13   \n",
            "   2    |  1720   |   0.011776   |     -      |     -     |   7.16   \n",
            "   2    |  1740   |   0.011864   |     -      |     -     |   7.17   \n",
            "   2    |  1760   |   0.022244   |     -      |     -     |   7.17   \n",
            "   2    |  1780   |   0.031848   |     -      |     -     |   7.12   \n",
            "   2    |  1800   |   0.011923   |     -      |     -     |   7.14   \n",
            "   2    |  1820   |   0.011771   |     -      |     -     |   7.14   \n",
            "   2    |  1840   |   0.001813   |     -      |     -     |   7.13   \n",
            "   2    |  1860   |   0.011838   |     -      |     -     |   7.15   \n",
            "   2    |  1880   |   0.001536   |     -      |     -     |   7.13   \n",
            "   2    |  1900   |   0.011618   |     -      |     -     |   7.13   \n",
            "   2    |  1920   |   0.011622   |     -      |     -     |   7.13   \n",
            "   2    |  1940   |   0.011515   |     -      |     -     |   7.11   \n",
            "   2    |  1960   |   0.001404   |     -      |     -     |   7.13   \n",
            "   2    |  1980   |   0.022320   |     -      |     -     |   7.14   \n",
            "   2    |  2000   |   0.001418   |     -      |     -     |   7.13   \n",
            "   2    |  2020   |   0.031833   |     -      |     -     |   7.15   \n",
            "   2    |  2040   |   0.001510   |     -      |     -     |   7.11   \n",
            "   2    |  2060   |   0.020948   |     -      |     -     |   7.15   \n",
            "   2    |  2080   |   0.001475   |     -      |     -     |   7.15   \n",
            "   2    |  2100   |   0.011697   |     -      |     -     |   7.18   \n",
            "   2    |  2120   |   0.011418   |     -      |     -     |   7.13   \n",
            "   2    |  2140   |   0.021431   |     -      |     -     |   7.11   \n",
            "   2    |  2160   |   0.011406   |     -      |     -     |   7.12   \n",
            "   2    |  2180   |   0.012158   |     -      |     -     |   7.13   \n",
            "   2    |  2200   |   0.001625   |     -      |     -     |   7.13   \n",
            "   2    |  2220   |   0.011170   |     -      |     -     |   7.13   \n",
            "   2    |  2240   |   0.001407   |     -      |     -     |   7.19   \n",
            "   2    |  2260   |   0.011898   |     -      |     -     |   7.23   \n",
            "   2    |  2280   |   0.001344   |     -      |     -     |   7.26   \n",
            "   2    |  2300   |   0.011028   |     -      |     -     |   7.24   \n",
            "   2    |  2320   |   0.011883   |     -      |     -     |   7.23   \n",
            "   2    |  2340   |   0.001311   |     -      |     -     |   7.21   \n",
            "   2    |  2360   |   0.001275   |     -      |     -     |   7.14   \n",
            "   2    |  2380   |   0.011887   |     -      |     -     |   7.16   \n",
            "   2    |  2400   |   0.021607   |     -      |     -     |   7.13   \n",
            "   2    |  2420   |   0.011084   |     -      |     -     |   7.11   \n",
            "   2    |  2440   |   0.001435   |     -      |     -     |   7.12   \n",
            "   2    |  2460   |   0.033077   |     -      |     -     |   7.11   \n",
            "   2    |  2480   |   0.001452   |     -      |     -     |   7.12   \n",
            "   2    |  2500   |   0.001410   |     -      |     -     |   7.10   \n",
            "   2    |  2520   |   0.032008   |     -      |     -     |   7.11   \n",
            "   2    |  2540   |   0.001407   |     -      |     -     |   7.12   \n",
            "   2    |  2560   |   0.001414   |     -      |     -     |   7.11   \n",
            "   2    |  2580   |   0.020498   |     -      |     -     |   7.13   \n",
            "   2    |  2600   |   0.010746   |     -      |     -     |   7.12   \n",
            "   2    |  2620   |   0.010877   |     -      |     -     |   7.12   \n",
            "   2    |  2640   |   0.001398   |     -      |     -     |   7.15   \n",
            "   2    |  2660   |   0.001373   |     -      |     -     |   7.14   \n",
            "   2    |  2680   |   0.001357   |     -      |     -     |   7.13   \n",
            "   2    |  2700   |   0.001352   |     -      |     -     |   7.15   \n",
            "   2    |  2720   |   0.001334   |     -      |     -     |   7.13   \n",
            "   2    |  2740   |   0.001316   |     -      |     -     |   7.12   \n",
            "   2    |  2760   |   0.010173   |     -      |     -     |   7.11   \n",
            "   2    |  2780   |   0.010680   |     -      |     -     |   7.12   \n",
            "   2    |  2800   |   0.022033   |     -      |     -     |   7.10   \n",
            "   2    |  2812   |   0.001310   |     -      |     -     |   4.21   \n",
            "----------------------------------------------------------------------\n",
            "   2    |    -    |   0.009906   |  0.009587  |   99.87   |  1041.28 \n",
            "----------------------------------------------------------------------\n",
            "\n",
            "\n",
            "Training complete!\n"
          ]
        }
      ],
      "source": [
        "set_seed(42)    # Set seed for reproducibility\n",
        "bert_classifier, optimizer, scheduler = initialize_model(epochs=2)\n",
        "train(bert_classifier, train_dataloader, val_dataloader, epochs=2, evaluation=True)"
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "import torch.nn.functional as F\n",
        "\n",
        "def bert_predict(model, test_dataloader):\n",
        "    \"\"\"Perform a forward pass on the trained BERT model to predict probabilities\n",
        "    on the test set.\n",
        "    \"\"\"\n",
        "    # Put the model into the evaluation mode. The dropout layers are disabled during\n",
        "    # the test time.\n",
        "    model.eval()\n",
        "\n",
        "    all_logits = []\n",
        "\n",
        "    # For each batch in our test set...\n",
        "    for batch in test_dataloader:\n",
        "        # Load batch to GPU\n",
        "        b_input_ids, b_attn_mask = tuple(t.to(device) for t in batch)[:2]\n",
        "\n",
        "        # Compute logits\n",
        "        with torch.no_grad():\n",
        "            logits = model(b_input_ids, b_attn_mask)\n",
        "        all_logits.append(logits)\n",
        "\n",
        "    # Concatenate logits from each batch\n",
        "    all_logits = torch.cat(all_logits, dim=0)\n",
        "\n",
        "    # Apply softmax to calculate probabilities\n",
        "    probs = F.softmax(all_logits, dim=1).cpu().numpy()\n",
        "\n",
        "    return probs"
      ],
      "metadata": {
        "id": "sXSmYpn9VMgZ"
      },
      "execution_count": 60,
      "outputs": []
    },
    {
      "cell_type": "code",
      "execution_count": 61,
      "metadata": {
        "id": "i_3c0axz-E-1",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 509
        },
        "outputId": "05602631-faa6-4c22-a544-3ca852c507c0"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "AUC: 0.7675\n",
            "Accuracy: 99.87%\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 640x480 with 1 Axes>"
            ],
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAkIAAAHHCAYAAABTMjf2AAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjcuMSwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy/bCgiHAAAACXBIWXMAAA9hAAAPYQGoP6dpAABqK0lEQVR4nO3dd1yV5f/H8RcgS0XcG0XNmXvmzEGplWWZ4kjRHJkzTXOvhmbmquxrmoqW5Sgty9LSstQszZEzLUfuneACBK7fH/ePowgYB4Gb8X4+HucB93Wu+z6fc26RD9d0McYYRERERDIhV7sDEBEREbGLEiERERHJtJQIiYiISKalREhEREQyLSVCIiIikmkpERIREZFMS4mQiIiIZFpKhERERCTTUiIkIiIimZYSIZFk4u/vT9euXe0OI9Np3LgxjRs3tjuM/zR+/HhcXFy4ePGi3aGkOS4uLowfPz5ZrnXs2DFcXFwIDg5OlutJxqdESNKF4OBgXFxcHI8sWbJQpEgRunbtyqlTp+wOL027fv06r732GpUrVyZr1qz4+vrSsGFDFi1aRHrZYWf//v2MHz+eY8eO2R1KHFFRUSxYsIDGjRuTO3duPD098ff3p1u3bvz+++92h5csPvnkE2bMmGF3GLGkxZgkfcpidwAiznj11VcpUaIEYWFh/PrrrwQHB7Np0yb27t2Ll5eXrbEdPHgQV9e09bfFuXPnaNasGQcOHKB9+/b069ePsLAwPv/8c4KCgvjmm29YvHgxbm5udod6T/v372fChAk0btwYf3//WM9999139gQF3Lx5k2eeeYY1a9bQqFEjRo4cSe7cuTl27BjLli1j4cKFHD9+nKJFi9oWY3L45JNP2Lt3Ly+99FKKXP/mzZtkyeLcr6OEYipevDg3b97E3d09GSOUjEyJkKQrLVu2pGbNmgD06NGDvHnzMnnyZFatWkW7du1sjc3T0zPVXzMsLAwPD48EE7CgoCAOHDjAypUrefLJJx3lAwYMYOjQobz99ttUq1aNYcOGpVbIgNVKlS1btmS5loeHR7JcJymGDh3KmjVrmD59epxfyOPGjWP69OmpGo8xhrCwMLy9vVP1dZMiOjqaiIgIvLy8kvWPGBcXF9v/KJJ0xoikAwsWLDCA2bZtW6zyr7/+2gBm4sSJscoPHDhg2rRpY3LlymU8PT1NjRo1zJdffhnnuv/++6956aWXTPHixY2Hh4cpUqSI6dy5s7lw4YKjTlhYmBk7dqwpVaqU8fDwMEWLFjVDhw41YWFhsa5VvHhxExQUZIwxZtu2bQYwwcHBcV5zzZo1BjBfffWVo+zkyZOmW7duJn/+/MbDw8NUqFDBzJs3L9Z5P/74owHMp59+akaNGmUKFy5sXFxczL///hvvZ7ZlyxYDmOeffz7e52/dumVKly5tcuXKZW7cuGGMMebo0aMGMFOmTDHTpk0zxYoVM15eXqZRo0Zmz549ca6RmM855t5t2LDBvPjiiyZfvnwmZ86cxhhjjh07Zl588UVTpkwZ4+XlZXLnzm2effZZc/To0Tjn3/348ccfjTHGPPzww+bhhx+O8zktXbrUvP7666ZIkSLG09PTNG3a1Pz1119x3sN7771nSpQoYby8vEytWrXMzz//HOea8Tlx4oTJkiWLeeSRR+5ZL8a4ceMMYP766y8TFBRkfH19TY4cOUzXrl3N9evXY9WdP3++adKkicmXL5/x8PAw5cuXN++//36caxYvXtw8/vjjZs2aNaZGjRrG09PTTJ8+3alrGGPMN998Yxo1amSyZ89ufHx8TM2aNc3ixYuNMdbne/dnX7x4cce5if35AEzfvn3Nxx9/bCpUqGCyZMliVq5c6Xhu3LhxjrqhoaFm4MCBjp/LfPnymYCAALN9+/b/jCnm3/CCBQtivf6BAwdM27ZtTd68eY2Xl5cpU6aMGTly5L1umWQSahGSdC1mzEiuXLkcZfv27aN+/foUKVKE4cOHky1bNpYtW0br1q35/PPPefrppwG4du0aDRs25MCBAzz//PNUr16dixcvsmrVKk6ePEnevHmJjo7mySefZNOmTfTq1Yvy5cuzZ88epk+fzqFDh/jiiy/ijatmzZqULFmSZcuWERQUFOu5pUuXkitXLpo3bw5Y3VcPPfQQLi4u9OvXj3z58vHtt9/SvXt3QkND47Q0vPbaa3h4eDBkyBDCw8MTbBH56quvAOjSpUu8z2fJkoWOHTsyYcIENm/eTEBAgOO5RYsWcfXqVfr27UtYWBgzZ86kadOm7NmzhwIFCjj1Ocfo06cP+fLlY+zYsVy/fh2Abdu28csvv9C+fXuKFi3KsWPH+N///kfjxo3Zv38/WbNmpVGjRgwYMIB33nmHkSNHUr58eQDH14S8+eabuLq6MmTIEEJCQnjrrbfo1KkTv/32m6PO//73P/r160fDhg0ZNGgQx44do3Xr1uTKles/u7O+/fZbIiMj6dy58z3r3a1du3aUKFGCSZMmsWPHDj788EPy58/P5MmTY8X14IMP8uSTT5IlSxa++uor+vTpQ3R0NH379o11vYMHD9KhQwdeeOEFevbsSdmyZZ26RnBwMM8//zwPPvggI0aMIGfOnOzcuZM1a9bQsWNHRo0aRUhICCdPnnS0cGXPnh3A6Z+PH374gWXLltGvXz/y5s0bp5szRu/evfnss8/o168fFSpU4NKlS2zatIkDBw5QvXr1e8YUn927d9OwYUPc3d3p1asX/v7+HD58mK+++oo33ngjcTdOMi67MzGRxIhpFVi3bp25cOGCOXHihPnss89Mvnz5jKenpzlx4oSjbrNmzUylSpVi/UUaHR1t6tWrZ0qXLu0oGzt2rAHMihUr4rxedHS0McaYjz76yLi6upqNGzfGen727NkGMJs3b3aU3dkiZIwxI0aMMO7u7uby5cuOsvDwcJMzZ85YrTTdu3c3hQoVMhcvXoz1Gu3btze+vr6O1pqYlo6SJUs6yu6ldevWBkiwxcgYY1asWGEA88477xhjbv817e3tbU6ePOmo99tvvxnADBo0yFGW2M855t41aNDAREZGxnr9+N5HTEvWokWLHGXLly+P1Qp0p4RahMqXL2/Cw8Md5TNnzjSAo2UrPDzc5MmTx9SqVcvcunXLUS84ONgA/9kiNGjQIAOYnTt33rNejJgWobtb6J5++mmTJ0+eWGXxfS7Nmzc3JUuWjFVWvHhxA5g1a9bEqZ+Ya1y5csX4+PiYOnXqmJs3b8aqG/MzYIwxjz/+eKxWoBjO/HwAxtXV1ezbty/OdbirRcjX19f07ds3Tr07JRRTfC1CjRo1Mj4+Puaff/5J8D1K5pW2RnaK/IeAgADy5cuHn58fzz77LNmyZWPVqlWOv94vX77MDz/8QLt27bh69SoXL17k4sWLXLp0iebNm/PXX385Zpl9/vnnVKlSJU7LBVjjDACWL19O+fLlKVeunONaFy9epGnTpgD8+OOPCcYaGBjIrVu3WLFihaPsu+++48qVKwQGBgLWmI7PP/+cVq1aYYyJ9RrNmzcnJCSEHTt2xLpuUFBQosaAXL16FQAfH58E68Q8FxoaGqu8devWFClSxHFcu3Zt6tSpwzfffAM49znH6NmzZ5xB2Xe+j1u3bnHp0iUeeOABcubMGed9O6tbt26xWssaNmwIwJEjRwD4/fffuXTpEj179ow1ULdTp06xWhgTEvOZ3evzjU/v3r1jHTds2JBLly7Fugd3fi4hISFcvHiRhx9+mCNHjhASEhLr/BIlSjhaF++UmGt8//33XL16leHDh8cZVxPzM3Avzv58PPzww1SoUOE/r5szZ05+++03Tp8+/Z91/8uFCxf4+eefef755ylWrFis5xLzHiXjU9eYpCuzZs2iTJkyhISEMH/+fH7++edYg5T//vtvjDGMGTOGMWPGxHuN8+fPU6RIEQ4fPkybNm3u+Xp//fUXBw4cIF++fAleKyFVqlShXLlyLF26lO7duwNWt1jevHkdvyguXLjAlStXmDNnDnPmzEnUa5QoUeKeMceI+QV99epVcubMGW+dhJKl0qVLx6lbpkwZli1bBjj3Od8r7ps3bzJp0iQWLFjAqVOnYk3nv/sXvrPu/qUXk9z8+++/APzzzz8APPDAA7HqZcmSJcEumzvlyJEDuP0ZJkdcMdfcvHkz48aNY8uWLdy4cSNW/ZCQEHx9fR3HCf17SMw1Dh8+DEDFihWdeg8xnP35SOy/3bfeeougoCD8/PyoUaMGjz32GF26dKFkyZJOxxiT+Cb1PUrGp0RI0pXatWs7Zo21bt2aBg0a0LFjRw4ePEj27NmJjo4GYMiQIfH+lQxxf/HdS3R0NJUqVWLatGnxPu/n53fP8wMDA3njjTe4ePEiPj4+rFq1ig4dOjhaIGLife655+KMJYpRuXLlWMeJnRFUvnx5vvjiC3bv3k2jRo3irbN7926ARP2VfqekfM7xxd2/f38WLFjASy+9RN26dfH19cXFxYX27ds7XiOpEloSwCTT2knlypUDYM+ePVStWjXR5/1XXIcPH6ZZs2aUK1eOadOm4efnh4eHB9988w3Tp0+P87nE97k6e42kcvbnI7H/dtu1a0fDhg1ZuXIl3333HVOmTGHy5MmsWLGCli1b3nfcIndSIiTplpubG5MmTaJJkya89957DB8+3PEXo7u7e6zBv/EpVaoUe/fu/c86f/zxB82aNUtSM3pgYCATJkzg888/p0CBAoSGhtK+fXvH8/ny5cPHx4eoqKj/jNdZTzzxBJMmTWLRokXxJkJRUVF88skn5MqVi/r168d67q+//opT/9ChQ46WEmc+53v57LPPCAoKYurUqY6ysLAwrly5EqteSnRhFC9eHLBat5o0aeIoj4yM5NixY3ES0Lu1bNkSNzc3Pv74Y6cHTN/LV199RXh4OKtWrYrVenSvbtikXqNUqVIA7N27955/ICT0+d/vz8e9FCpUiD59+tCnTx/Onz9P9erVeeONNxyJUGJfL+bf6n/9rEvmpTFCkq41btyY2rVrM2PGDMLCwsifPz+NGzfmgw8+4MyZM3HqX7hwwfF9mzZt+OOPP1i5cmWcejF/nbdr145Tp04xd+7cOHVu3rzpmP2UkPLly1OpUiWWLl3K0qVLKVSoUKykxM3NjTZt2vD555/H+x/1nfE6q169egQEBLBgwQK+/vrrOM+PGjWKQ4cO8corr8T5S/2LL76INcZn69at/Pbbb45fQs58zvfi5uYWp4Xm3XffJSoqKlZZzJpDdydI96NmzZrkyZOHuXPnEhkZ6ShfvHixo/vsXvz8/OjZsyffffcd7777bpzno6OjmTp1KidPnnQqrpgWo7u7CRcsWJDs13j00Ufx8fFh0qRJhIWFxXruznOzZcsWb1fl/f58xCcqKirOa+XPn5/ChQsTHh7+nzHdLV++fDRq1Ij58+dz/PjxWM8lV+ugpG9qEZJ0b+jQobRt25bg4GB69+7NrFmzaNCgAZUqVaJnz56ULFmSc+fOsWXLFk6ePMkff/zhOO+zzz6jbdu2PP/889SoUYPLly+zatUqZs+eTZUqVejcuTPLli2jd+/e/Pjjj9SvX5+oqCj+/PNPli1bxtq1ax1ddQkJDAxk7NixeHl50b179ziLH7755pv8+OOP1KlTh549e1KhQgUuX77Mjh07WLduHZcvX07yZ7No0SKaNWvGU089RceOHWnYsCHh4eGsWLGCDRs2EBgYyNChQ+Oc98ADD9CgQQNefPFFwsPDmTFjBnny5OGVV15x1Ens53wvTzzxBB999BG+vr5UqFCBLVu2sG7dOvLkyROrXtWqVXFzc2Py5MmEhITg6elJ06ZNyZ8/f5I/Gw8PD8aPH0///v1p2rQp7dq149ixYwQHB1OqVKlEtThMnTqVw4cPM2DAAFasWMETTzxBrly5OH78OMuXL+fPP/+M1QKYGI8++igeHh60atWKF154gWvXrjF37lzy588fb9J5P9fIkSMH06dPp0ePHtSqVYuOHTuSK1cu/vjjD27cuMHChQsBqFGjBkuXLmXw4MHUqlWL7Nmz06pVq2T5+bjb1atXKVq0KM8++yxVqlQhe/bsrFu3jm3btsVqOUwopvi88847NGjQgOrVq9OrVy9KlCjBsWPHWL16Nbt27XIqPsmAbJmrJuKkhBZUNMaYqKgoU6pUKVOqVCnH9OzDhw+bLl26mIIFCxp3d3dTpEgR88QTT5jPPvss1rmXLl0y/fr1M0WKFHEsBhcUFBRrKntERISZPHmyefDBB42np6fJlSuXqVGjhpkwYYIJCQlx1Lt7+nyMv/76y7Ho26ZNm+J9f+fOnTN9+/Y1fn5+xt3d3RQsWNA0a9bMzJkzx1EnZlr48uXLnfrsrl69asaPH28efPBB4+3tbXx8fEz9+vVNcHBwnOnDdy6oOHXqVOPn52c8PT1Nw4YNzR9//BHn2on5nO917/7991/TrVs3kzdvXpM9e3bTvHlz8+eff8b7Wc6dO9eULFnSuLm5JWpBxbs/p4QW2nvnnXdM8eLFjaenp6ldu7bZvHmzqVGjhmnRokUiPl1jIiMjzYcffmgaNmxofH19jbu7uylevLjp1q1brKn1MdPn71ys887P585FJFetWmUqV65svLy8jL+/v5k8ebKZP39+nHoxCyrGJ7HXiKlbr1494+3tbXLkyGFq165tPv30U8fz165dMx07djQ5c+aMs6BiYn8++P8FFePDHdPnw8PDzdChQ02VKlWMj4+PyZYtm6lSpUqcxSATiimh+7x3717z9NNPm5w5cxovLy9TtmxZM2bMmHjjkczFxRi1DYqI5dixY5QoUYIpU6YwZMgQu8OxRXR0NPny5eOZZ56Jt8tHRDIWjRESkUwrLCwszjiRRYsWcfnyZRo3bmxPUCKSqjRGSEQyrV9//ZVBgwbRtm1b8uTJw44dO5g3bx4VK1akbdu2docnIqlAiZCIZFr+/v74+fnxzjvvcPnyZXLnzk2XLl148803bd3VXkRSj61jhH7++WemTJnC9u3bOXPmDCtXrqR169b3PGfDhg0MHjyYffv24efnx+jRo+natWuqxCsiIiIZi61jhK5fv06VKlWYNWtWouofPXqUxx9/nCZNmrBr1y5eeuklevTowdq1a1M4UhEREcmI0sysMRcXl/9sERo2bBirV6+OtfBc+/btuXLlCmvWrEmFKEVERCQjSVdjhLZs2RJnOf/mzZvz0ksvJXhOeHh4rNVIo6OjuXz5Mnny5NHOwyIiIumEMYarV69SuHDhOAvT3o90lQidPXuWAgUKxCqL2b/p5s2b8W7oN2nSJCZMmJBaIYqIiEgKOnHiBEWLFk2266WrRCgpRowYweDBgx3HISEhFCtWjBMnTpAjRw4bIxMRkczs5EnYt8/uKOJasADOn4dkbHRJsoaXv2RrzkcId83Kvn2h3Ljhh4+PT7K+RrpKhAoWLMi5c+dilZ07d44cOXLE2xoE4OnpiaenZ5zyHDlyKBESEUkm16/DlClw13/RtouMhK+/hgoV7I4ktvBw2LzZ7ijSrqxcZxZ96cpC5tKDXtxe5T25h7Wkq0Sobt26fPPNN7HKvv/+e+rWrWtTRCIiAlaykZZHIZw9a3cECatRA9LakNVr12DiRHtahXz+2UutKe3wOXkA4+pKg8BifNHOcOMmdOyY/K9nayJ07do1/v77b8fx0aNH2bVrF7lz56ZYsWKMGDGCU6dOsWjRIgB69+7Ne++9xyuvvMLzzz/PDz/8wLJly1i9erVdb0FERIAbN6yvpUtDp072xnK3qCgwJu21CgHUrg2lStkdRRphDMyfD8P7w82bUKgQLp98QvnGjSkPhIamzMvamgj9/vvvNGnSxHEcM5YnKCiI4OBgzpw5w/Hjxx3PlyhRgtWrVzNo0CBmzpxJ0aJF+fDDD2nevHmqxy4iYodr12DJErh61e5IYluwwPpapgyMG2dvLJIOXbsGvXvD4sXW8aOPwkcfQf78Kf7SaWYdodQSGhqKr68vISEhGiMkIunOW2/BsGF2R5GwNm3gs8/sjkLSnZMnoWpVuHIFXn8dXnklTr9cSv3+TldjhEREMrvLl62v5cpZY0vSEg8P6N/f7igkXSpaFD79FLy9oUGDVH1pJUIiIunQY4/B1Kl2RyGSRKGh0KsXtG8PMTtKPPKILaGkgVUCREREJNPYvh2qV4elS+GFF26PtLeJEiERERFJecbAu+9CvXpw+DAULw5ffglZs9oalrrGREREJGVduQLdu8OKFdZx69bWVPlcueyMClAiJCIiIinpyhWoVg2OHQN3d3j7bWtUfRpZRVJdYyIiIpJycuaEli2hZEn45RcYMCDNJEGgFiERERFJbpcuWRu9FShgHU+bZm2w5utrb1zxUIuQiIiIJJ9ffrG6wjp0sPY3AfDySpNJECgREhERkeQQHQ2TJ0OjRnDihPU4c8buqP6TEiERERG5PxcuwBNPwPDhVitQhw6wY4e1YnQapzFCIiIiknQbN1orRJ8+bXWBvfMO9OiRpgZE34sSIREREUmaqCjo08dKgsqVg2XLoFIlu6NyirrGREREJGnc3KzNUnv0gG3b0l0SBEqERERExBk//AAffHD7uGJFmDsXsme3L6b7oERIRERE/ltUFIwbBwEB0K8f/P673RElC40REhERkXs7fRo6dYING6zjrl2hQgU7I0o2SoREREQkYWvXQufO1hT57NmtbrGOHe2OKtmoa0xERETiN348tGhhJUFVqsD27RkqCQIlQiIiIpKQnDmtr717w6+/QpkytoaTEtQ1JiIiIrddvw7ZslnfDxxo7Rv28MP2xpSC1CIkIiIicOsWDB0K1avD1atWmYtLhk6CQImQiIiI/POPtVnq22/DoUPwxRd2R5RqlAiJiIhkZl9+CVWrWmOAfH3h88+tWWKZhBIhERGRzCgiAl56CVq3hitXoHZt2LkTnnnG5sBSlxIhERGRzGjYMJg50/r+5ZetXeRLlLA3JhsoERIREcmMhg+HBx+EVaussUEeHnZHZAslQiIiIplBWJi1U3yMAgVg925o1cq+mNIArSMkIiKS0f31F7RrB7t2WccdOlhfXdUeok9AREQkI/v0U2ttoF27IG9eyJ3b7ojSFCVCIiIiGdHNm9Crl7U32LVr1jpBu3ZB8+Z2R5amKBESERHJaP78E+rUgblzrdWhR4+G9euhSBG7I0tzNEZIREQkozl8GPbsgfz5YfFiCAiwO6I0S4mQiEg81q6FceOsNefSklOn7I5A0oXHH7dagx5/HAoVsjuaNE2JkIhIPD74AH77ze4oElasmN0RSJqybx/07g0ffwzFi1tlPXrYG1M6oURIRGw1eTJMnQrR0bfLjIldJ7mPE1Pn5k3r64AB0LJl3PPt5OMDdevaHYWkCcbAggXQr5/1j/all2DlSrujSleUCImIrYKD4cIFu6OIn7u7tdzKQw/ZHYlIPK5ds1qBFi+2jh991GrKFKcoERKRNCE4GGrVun3s4hL7eWeOk6tuzpyQK1dCEYvY6I8/rAUSDx0CNzd47TVr7zAtkOg0JUIimcTVq7B/v91RxBXTBVWiBFSoYG8sIunCxo3wyCMQHm5Nh1+yBBo0sDuqdEuJkEgyMgaGDr29in1asn693RHc290tMyKSgFq1oFw5KwlauNBaLVqSTImQSDI6etQa+JuWFSgAWbPaHUVsJUtCzZp2RyGShh04AGXKWN1gXl6wbp21VYa6wu6bEiGRZBQZaX3NmhU+/NDeWOJTpYq6n0TSFWNg1ix4+WUYNQrGjrXK1QqUbJQIiaQAD4/bmzuLiCTJlSvQvTusWGEd//GHtc6EWoGSlT5NERGRtGbrVqhWzUqC3N1hxgz47DMlQSlAn6iIiEhaYQxMn27NAjt2zJpOuXkzDByoGQUpRImQiIhIWnH0KIwcCbduQZs2sGNH7AW2JNlpjJCIiEhaUbKkNTj65k3o00etQKlAiZCIiIhdoqOtNTcaNry9l8vzz9sbUyajREhERMQOFy5AUBB8+621Y/zevZA9u91RZTpKhERERFLbzz9ba2ycPm0tkDhqFGTLZndUmZIGS4uIiKSW6Gh44w1o0sRKgsqWhd9+g549NR7IJmoREhERSQ3XrsEzz8D331vHnTvD+++rO8xmSoRERERSQ7Zs4O1tPd5/H7p2tTsiQYmQiIhIyomKgogIK/lxcYEFC+DsWW36l4ZojJCIiEhKOHMGAgKs8T/GWGW5cysJSmPUIiQiIpLcvvsOnnvOmiKfLRscOQKlStkdlcRDLUIiIiLJJTLSmgrfooWVBFWuDL//riQoDVOLkIiISHI4eRI6doSNG63jF16wNlD19rY3LrknJUIiIiL3KzoaWra0Vof28YG5cyEw0O6oJBHUNSYiInK/XF1hxgyoWdPaMV5JULqhREhERCQpjh+3BkXHaNbMWiX6gQfsi0mcpkRIRETEWatWQdWq8Oyz8Pfft8td9Ws1vdEdExERSayICBg0CJ56Cv79F8qVgywabpue2Z4IzZo1C39/f7y8vKhTpw5bt269Z/0ZM2ZQtmxZvL298fPzY9CgQYSFhaVStCIikmkdPQoNGlhjgcBKiDZtAn9/O6OS+2RrGrt06VIGDx7M7NmzqVOnDjNmzKB58+YcPHiQ/Pnzx6n/ySefMHz4cObPn0+9evU4dOgQXbt2xcXFhWnTptnwDsROoaHWMh1RUcn7iI5O+rkXL9r9qYhIivj8c+jeHUJCIFcuCA6GJ5+0OypJBrYmQtOmTaNnz55069YNgNmzZ7N69Wrmz5/P8OHD49T/5ZdfqF+/Ph07dgTA39+fDh068Ntvv6Vq3GK/w4ehUiW4edPuSOKXI4fdEYhIsvrlFysJqlsXliyBYsXsjkiSiW2JUEREBNu3b2fEiBGOMldXVwICAtiyZUu859SrV4+PP/6YrVu3Urt2bY4cOcI333xD586dE3yd8PBwwsPDHcehoaHJ9ybENvv3W0mQq6u1ZIeb2/0/XF2T5zpubvD443Z/QiJy34yxNkoFmDQJiheHF18Ed3d745JkZVsidPHiRaKioihQoECs8gIFCvDnn3/Ge07Hjh25ePEiDRo0wBhDZGQkvXv3ZuTIkQm+zqRJk5gwYUKyxi5pR61a8OuvdkchIhnOkiWwcKE1O8zdHTw8YMAAu6OSFGD7YGlnbNiwgYkTJ/L++++zY8cOVqxYwerVq3nttdcSPGfEiBGEhIQ4HidOnEjFiEVEJF25edPaGqNDB1izxlohWjI021qE8ubNi5ubG+fOnYtVfu7cOQoWLBjvOWPGjKFz58706NEDgEqVKnH9+nV69erFqFGjcI1n/QZPT088PT2T/w2IiEjGcvAgtGsHu3dbXWIjR0KvXnZHJSnMthYhDw8PatSowfr16x1l0dHRrF+/nrp168Z7zo0bN+IkO25ubgAYY1IuWBERydg+/hhq1LCSoPz5Ye1aeP11rRGUCdh6hwcPHkxQUBA1a9akdu3azJgxg+vXrztmkXXp0oUiRYowadIkAFq1asW0adOoVq0aderU4e+//2bMmDG0atXKkRCJiIg45Y03YPRo6/smTWDxYihUyN6YJNXYmggFBgZy4cIFxo4dy9mzZ6latSpr1qxxDKA+fvx4rBag0aNH4+LiwujRozl16hT58uWjVatWvPHGG3a9BRERSe+efRbeegsGD7YSIv1hnam4mEzWpxQaGoqvry8hISHk0GIv6dZXX1lrmdWpo1ljIuIkY6wusCpVbpddugR58tgXk/ynlPr9na5mjYmIiNyXa9egSxeoXh1++ul2uZKgTEuJkIiIZA67d0PNmtbAaIC9e+2NR9IEJUIiIpKxGQNz5kDt2tYU+SJFYMMG6NvX7sgkDdC8QBERybhCQ60FEpcssY5btoRFiyBvXnvjkjRDLUIiIpJxffmllQS5uVkzw77+WkmQxKIWIRERybieew527oS2ba2d40XuohYhERHJOK5cgX794N9/rWMXF5g2TUmQJEgtQiIikjFs2waBgXD0KFy8eHtckMg9qEVIRETSN2NgxgyoX99KgkqUgJdftjsqSSfUIiQiIunX5cvQrRusWmUdt2kDH34IOXPaGpakH0qEREQkfdqzB554Ao4fBw8PayxQnz7WuCCRRFIiJCIi6VPhwla3WKlSsGyZtW2GiJOUCImISPpx9Spkz261+uTJA99+C35+oE20JYk0WFpERNKHjRuhfHkIDr5d9uCDSoLkvigREhGRtC06GiZOhCZN4NQpePddiIqyOyrJIJQIiYhI2nX+PLRoAaNGWcnPc8/Bzz9bW2aIJAONERIRkbTpxx+hY0c4exa8veG996yp8poVJslIiZCIiKQ9//wDjz4KkZFQoYI1K+zBB+2OSjIgJUIiIpL2FC8OI0bAyZPWmKBs2eyOSDIoJUIiIpI2rFsH/v7wwAPW8YQJ6gaTFKfB0iIiYq/ISBg92uoKCwyE8HCrXEmQpAK1CImIiH1OnYIOHaw1ggBq1bJWixZJJUqERETEHt9+C126wMWL4OMDc+ZA+/Z2RyWZjLrGREQkdd26BcOGwWOPWUlQtWqwfbuSILGFEiEREUldxlhrBAH07Qu//AKlS9sbk2Ra6hoTEZHUYYw1ANrDA5YuhR07oE0bu6OSTE6JkIiIpKyICBg+HLy8rD3DAEqUsB4iNlMiJCIiKefoUWvsz9atVmtQly5QrpzdUYk4aIyQiIikjBUrrIHQW7dCzpywcqWSIElzlAiJiEjyCg+H/v2t8T8hIfDQQ7BrFzz1lN2RicShrjEREUk+xlgrRP/8s3X8yivw+uvg7m5vXCIJUCIkIiLJx8UFevSAfftg0SJrrSCRNExdYyIicn9u3oQDB24fd+4Mhw4pCZJ0QYmQiIgk3cGD1higgAC4cOF2ee7c9sUk4gQlQiIikjQffww1asDu3da2GUeP2h2RiNOUCImIiHNu3IDu3a0usOvXoXFja1ZY7dp2RybiNCVCIiKSePv3WwnP/PnWwOhx42DdOihc2O7IRJJEs8ZERCTxJk+2ZoQVLAiLF0PTpnZHJHJflAiJiEjivfMOZMli7RlWoIDd0YjcN3WNiYhIwvbsgaFDrYUSAXx9Yd48JUGSYahFSERE4jIGPvwQBgyAsDAoW9ZaKFEkg1EiJCIisYWGwgsvwJIl1nHLltonTDIsdY2JiMhtO3daawMtWQJubtbg6K+/hnz57I5MJEXcV4tQWFgYXl5eyRWLiIjY6aOPrO6viAjw87OSoXr17I5KJEU53SIUHR3Na6+9RpEiRciePTtHjhwBYMyYMcybNy/ZAxQRkVRSogRERUGrVtYCiUqCJBNwOhF6/fXXCQ4O5q233sLDw8NRXrFiRT788MNkDU5ERFJYSMjt7xs0gC1b4MsvtVeYZBpOJ0KLFi1izpw5dOrUCTc3N0d5lSpV+PPPP5M1OBERSSHGwMyZ4O9vrRYdo1Yta8VokUzC6UTo1KlTPPDAA3HKo6OjuXXrVrIEJSIiKejyZXj6aXjpJbhyBYKDbQ5IxD5OJ0IVKlRg48aNcco/++wzqlWrlixBiYhICvn1V6hWzer+8vCAd9+1ZoaJZFJOzxobO3YsQUFBnDp1iujoaFasWMHBgwdZtGgRX3/9dUrEKCIi9ys6GqZNgxEjIDISSpWCpUutqfIimZjTLUJPPfUUX331FevWrSNbtmyMHTuWAwcO8NVXX/HII4+kRIwiInK/Pv7Y2iojMhLatYPt25UEiZDEdYQaNmzI999/n9yxiIhISunY0dot/umnrVWjNSBaBEhCi1DJkiW5dOlSnPIrV65QsmTJZAlKRETuU3S0tVdYeLh1nCULrFkDvXsrCRK5g9OJ0LFjx4iKiopTHh4ezqlTp5IlKBERuQ/nz1v7g/XsCcOG3S5XAiQSR6K7xlatWuX4fu3atfj6+jqOo6KiWL9+Pf7+/skanIiIOGnDBqsb7MwZ8PaGypXtjkgkTUt0ItS6dWsAXFxcCAoKivWcu7s7/v7+TJ06NVmDExGRRIqKgjfegAkTrG6x8uVh+XJ48EG7IxNJ0xKdCEVHRwNQokQJtm3bRt68eVMsKBERccLZs9CpE/zwg3XcrZu1PlC2bPbGJZIOOD1r7OjRoykRh4iIJNWNG/D775A1K8yeDZ072x2RSLqRpOnz169f56effuL48eNERETEem7AgAHJEpiIiNyDMbcHP5csCcuWQfHiUK6cvXGJpDNOJ0I7d+7kscce48aNG1y/fp3cuXNz8eJFsmbNSv78+ZUIiYiktFOn4LnnrFWiH33UKmve3N6YRNIpp6fPDxo0iFatWvHvv//i7e3Nr7/+yj///EONGjV4++23UyJGERGJsWYNVK1qzQ7r08daKVpEkszpRGjXrl28/PLLuLq64ubmRnh4OH5+frz11luMHDkyJWIUEZFbt2D4cGt9oIsXrWTom2+shRJFJMmcToTc3d1xdbVOy58/P8ePHwfA19eXEydOJG90IiICJ05A48a3d4nv0we2bIEyZWwNSyQjcDoRqlatGtu2bQPg4YcfZuzYsSxevJiXXnqJihUrOh3ArFmz8Pf3x8vLizp16rB169Z71r9y5Qp9+/alUKFCeHp6UqZMGb755hunX1dEJF04dcpq/fnlF8iRw1obaNYs8PKyOzKRDMHpRGjixIkUKlQIgDfeeINcuXLx4osvcuHCBT744AOnrrV06VIGDx7MuHHj2LFjB1WqVKF58+acP38+3voRERE88sgjHDt2jM8++4yDBw8yd+5cihQp4uzbEBFJH4oUgVatoGZN2LkTnn3W7ohEMhQXY4yx68Xr1KlDrVq1eO+99wBr0UY/Pz/69+/P8OHD49SfPXs2U6ZM4c8//8Td3T1JrxkaGoqvry8hISHkyJHjvuIX+3z1FTz5JNSpA7/+anc0Isns2DHInh1iFq69cQPc3MDT09awROyUUr+/nW4RSsiOHTt44oknEl0/IiKC7du3ExAQcDsYV1cCAgLYsmVLvOesWrWKunXr0rdvXwoUKEDFihWZOHFivJvAxggPDyc0NDTWQ0QkzVq50uoKCwqytsoAa6FEJUEiKcKpRGjt2rUMGTKEkSNHcuTIEQD+/PNPWrduTa1atRzbcCTGxYsXiYqKokCBArHKCxQowNmzZ+M958iRI3z22WdERUXxzTffMGbMGKZOncrrr7+e4OtMmjQJX19fx8PPzy/RMYqIpJrwcBgwAJ55BkJC4NIl66uIpKhEJ0Lz5s2jZcuWBAcHM3nyZB566CE+/vhj6tatS8GCBdm7d2+KD1qOjo4mf/78zJkzhxo1ahAYGMioUaOYPXt2gueMGDGCkJAQx0Mz20QkzTl8GOrXt/YHAxgyBDZuhFy57I1LJBNI9AIUM2fOZPLkyQwdOpTPP/+ctm3b8v7777Nnzx6KFi3q9AvnzZsXNzc3zp07F6v83LlzFCxYMN5zChUqhLu7O25ubo6y8uXLc/bsWSIiIvDw8IhzjqenJ55qUhaRtGrZMujRA65ehTx5YOFCePxxu6MSyTQS3SJ0+PBh2rZtC8AzzzxDlixZmDJlSpKSIAAPDw9q1KjB+vXrHWXR0dGsX7+eunXrxntO/fr1+fvvv2N1wR06dIhChQrFmwSJiKRpYWHWNhlXr1otQrt2KQkSSWWJToRu3rxJ1qxZAXBxccHT09MxjT6pBg8ezNy5c1m4cCEHDhzgxRdf5Pr163Tr1g2ALl26MGLECEf9F198kcuXLzNw4EAOHTrE6tWrmThxIn379r2vOEREbOHlBUuXwsiR1pYZSfzDUkSSzqm12T/88EOyZ88OQGRkJMHBweSNmd75/5zZdDUwMJALFy4wduxYzp49S9WqVVmzZo1jAPXx48cdq1gD+Pn5sXbtWgYNGkTlypUpUqQIAwcOZNiwYc68DRER+3zyiTUdvkcP67hmTeshIrZI9DpC/v7+uLi43PtiLi6O2WRpldYRyhi0jpCkOzduwMCB8OGH4OFhdYOVL293VCLpRkr9/k50i9CxY8eS7UVFRDKVAwegXTvYuxdcXKxxQdonTCRN0LbFIiIpaeFCa5PUGzegQAGra6xpU7ujEpH/p0RIRCQlGAM9e8K8edZxQAB8/LGVDIlImpFsW2yIiMgdXFygZElwdYXXXoM1a5QEiaRBahESEUkuxljbYuTMaR0PHw4tWkD16raGJSIJU4uQiEhyuHoVOnWChg2t8UBgtQYpCRJJ05KUCB0+fJjRo0fToUMHzp8/D8C3337Lvn37kjU4EZF0YdcuqFEDPv3UmiH28892RyQiieR0IvTTTz9RqVIlfvvtN1asWMG1a9cA+OOPPxg3blyyBygikmYZA//7Hzz0EPz1F/j5WUlQixZ2RyYiieR0IjR8+HBef/11vv/++1j7ezVt2pRftbKdiGQWISEQGGhNjQ8Ph1atYOdOqFfP7shExAlOJ0J79uzh6aefjlOeP39+Ll68mCxBiYikef36wfLlkCULTJ0KX35p7R4vIumK04lQzpw5OXPmTJzynTt3UqRIkWQJSkQkzZs0yRoXtGkTDB5sTZcXkXTH6USoffv2DBs2jLNnz+Li4kJ0dDSbN29myJAhdOnSJSViFBGx37//WqtExyhaFLZtsza8E5F0y+lEaOLEiZQrVw4/Pz+uXbtGhQoVaNSoEfXq1WP06NEpEaOIiL1++w2qVYOuXa0usBhqBRJJ95xeUNHDw4O5c+cyZswY9u7dy7Vr16hWrRqlS5dOifhEROxjDEybZi2MGBkJpUpZLUEikmE4nQht2rSJBg0aUKxYMYoVK5YSMYmI2O/SJasF6OuvreN27WDuXMiRw9awRCR5Od011rRpU0qUKMHIkSPZv39/SsQkImKvzZuhalUrCfL0tNYKWrJESZBIBuR0InT69GlefvllfvrpJypWrEjVqlWZMmUKJ0+eTIn4RERS3+nTcPIklC4Nv/4KvXtrPJBIBuV0IpQ3b1769evH5s2bOXz4MG3btmXhwoX4+/vTtGnTlIhRRCTlGXP7+7ZtITgYtm+3WoZEJMO6r01XS5QowfDhw3nzzTepVKkSP/30U3LFJSKSen76yVoT6M410oKCwMfHvphEJFUkORHavHkzffr0oVChQnTs2JGKFSuyevXq5IxNRCRlRUXBa69B06bW9hhjx9odkYikMqdnjY0YMYIlS5Zw+vRpHnnkEWbOnMlTTz1F1qxZUyI+EZGUcfYsPPccrF9vHXftCjNm2BmRiNjA6UTo559/ZujQobRr1468efOmREwiIilr/Xro1AnOnYOsWa1ZYVoZXyRTcjoR2rx5c0rEISKSOlauhDZtrMHRFSvCsmVQvrzdUYmITRKVCK1atYqWLVvi7u7OqlWr7ln3ySefTJbARERSxCOPQNmy0LAhzJwJ3t52RyQiNkpUItS6dWvOnj1L/vz5ad26dYL1XFxciIqKSq7YRESSx7Zt1qwwV1fInt1aG8jX1+6oRCQNSNSssejoaPLnz+/4PqGHkiARSVMiI2HECKhd29ozLIaSIBH5f05Pn1+0aBHh4eFxyiMiIli0aFGyBCUict9OnIDGjeHNN61jrX4vIvFwOhHq1q0bISEhccqvXr1Kt27dkiUoEZH7snq1tSL05s3W/mDLl2tqvIjEy+lEyBiDSzx77pw8eRJfNTeLiJ0iImDIEHjiCbh8GWrWtBZKfPZZuyMTkTQq0dPnq1WrhouLCy4uLjRr1owsWW6fGhUVxdGjR2nRokWKBCkikigHDsA771jfDxwIkydbu8eLiCQg0YlQzGyxXbt20bx5c7Jnz+54zsPDA39/f9q0aZPsAYqIJFqVKvDee5A/P9xjhquISIxEJ0Ljxo0DwN/fn8DAQLy8vFIsKBGRRAkPh5EjoXPn27vE9+pla0gikr44vbJ0UFBQSsQhIuKcw4chMBC2b4evv4a9e8Hd3e6oRCSdSVQilDt3bg4dOkTevHnJlStXvIOlY1y+fDnZghMRidfy5dCjB4SGQu7c1hpBSoJEJAkSlQhNnz4dHx8fx/f3SoRERFJMWBgMHmxtkgpQvz58+in4+dkbl4ikW4lKhO7sDuvatWtKxSIikrALF+DRR2HXLut4xAh49VXI4nQPv4iIg9PrCO3YsYM9e/Y4jr/88ktat27NyJEjiYiISNbgREQccueGvHkhXz5YswYmTlQSJCL3zelE6IUXXuDQoUMAHDlyhMDAQLJmzcry5ct55ZVXkj1AEcnEbtyAmzet793cYPFiq0WoeXNbwxKRjMPpROjQoUNU/f9pqsuXL+fhhx/mk08+ITg4mM8//zy54xORzOrAAahTB1566XZZ/vxQuLBtIYlIxpOkLTaio6MBWLduHY899hgAfn5+XLx4MXmjE5HMaeFCa3uMvXvhyy+t8UEiIinA6USoZs2avP7663z00Uf89NNPPP744wAcPXqUAgUKJHuAIpKJXL8OXbtajxs3oFkzqyssXz6bAxORjMrpRGjGjBns2LGDfv36MWrUKB544AEAPvvsM+rVq5fsAYpIJrF3L9SqZbUGubrCa6/B2rVQsKDdkYlIBub0lIvKlSvHmjUWY8qUKbi5uSVLUCKSyUREQMuWcPKkNQbok0/g4YftjkpEMoEkzz3dvn07Bw4cAKBChQpUr1492YISkUzGwwNmz4ZZs6wWIXWFiUgqcToROn/+PIGBgfz000/kzJkTgCtXrtCkSROWLFlCPv0HJiKJ8ccfcP48PPKIdfz44/DYY6CV60UkFTk9Rqh///5cu3aNffv2cfnyZS5fvszevXsJDQ1lwIABKRGjiGQkxlitP3XqWJumHj9++zklQSKSypxuEVqzZg3r1q2jfPnyjrIKFSowa9YsHn300WQNTkQymJAQ6NULli2zjh95BLJlszcmEcnUnG4Rio6Oxj2eXZ7d3d0d6wuJiMSxfTtUr24lQVmywNSpsGoV5Mljd2Qikok5nQg1bdqUgQMHcvr0aUfZqVOnGDRoEM2aNUvW4EQkg3j3XahXD44cgeLFYdMmaxd5dYWJiM2cToTee+89QkND8ff3p1SpUpQqVYoSJUoQGhrKu+++mxIxikh6t2+fNUW+dWvYudMaHyQikgY4PUbIz8+PHTt2sH79esf0+fLlyxMQEJDswYlIOmbM7Raf6dOtFqHOndUKJCJpilOJ0NKlS1m1ahURERE0a9aM/v37p1RcIpJeGWMlPt9/D19/be0a7+0NXbrYHZmISByJToT+97//0bdvX0qXLo23tzcrVqzg8OHDTJkyJSXjE5H05NIla5+wr7+2jlesgLZtbQ1JROReEj1G6L333mPcuHEcPHiQXbt2sXDhQt5///2UjE1E0pNffoFq1awkyNMT/vc/ePZZu6MSEbmnRCdCR44cISgoyHHcsWNHIiMjOXPmTIoEJiLpRHQ0TJ4MjRrBiRNQujT8+iv07q3xQCKS5iU6EQoPDyfbHQufubq64uHhwc2bN1MkMBFJJwYMgOHDISoKOna01guqWtXuqEREEsWpwdJjxowha9asjuOIiAjeeOMNfH19HWXTpk1LvuhEJO3r1Qs+/RTeeguef16tQCKSriQ6EWrUqBEHDx6MVVavXj2OHDniOHbRf4AiGV9UFPz+++21gCpXhmPHwMfH1rBERJIi0YnQhg0bUjAMEUkXzp2D556DDRus1aFjkiElQSKSTjm9srSIZFI//ABVqsC6deDhASdP2h2RiMh9UyIkIvcWFQXjxkFAgNUiVLGi1TXWpo3dkYmI3Dent9gQkUzk9Gno1MnqCgPo0QNmzoQ7Jk2IiKRnSoREJGErVlhJUPbs8MEH1vR4EZEMJE10jc2aNQt/f3+8vLyoU6cOW7duTdR5S5YswcXFhdatW6dsgCKZVd++MGSItTaQkiARyYCSlAht3LiR5557jrp163Lq1CkAPvroIzZt2uT0tZYuXcrgwYMZN24cO3bsoEqVKjRv3pzz58/f87xjx44xZMgQGjZsmJS3ICLxOXnS2ivs6lXr2MUFpkyBMmVsDUtEJKU4nQh9/vnnNG/eHG9vb3bu3El4eDgAISEhTJw40ekApk2bRs+ePenWrRsVKlRg9uzZZM2alfnz5yd4TlRUFJ06dWLChAmULFnS6dcUkXisXm2tCL1wIbz8st3RiIikCqcToddff53Zs2czd+5c3N3dHeX169dnx44dTl0rIiKC7du3ExAQcDsgV1cCAgLYsmVLgue9+uqr5M+fn+7du//na4SHhxMaGhrrISJ3uHULhg6FJ56wdo+vUQOGDbM7KhGRVOF0InTw4EEaNWoUp9zX15crV644da2LFy8SFRVFgQIFYpUXKFCAs2fPxnvOpk2bmDdvHnPnzk3Ua0yaNAlfX1/Hw8/Pz6kYRTK0f/6xNkt9+23reMAA2LwZSpWyNy4RkVTidCJUsGBB/v777zjlmzZtSvFuqqtXr9K5c2fmzp1L3rx5E3XOiBEjCAkJcTxOnDiRojGKpBsbN1pdYb/+CjlzwsqV1tR4T0+7IxMRSTVOT5/v2bMnAwcOZP78+bi4uHD69Gm2bNnCkCFDGDNmjFPXyps3L25ubpw7dy5W+blz5yhYsGCc+ocPH+bYsWO0atXKURYdHW29kSxZOHjwIKXu+kvW09MTT/3HLhJX6dJW0lOnDixZAv7+dkckIpLqnE6Ehg8fTnR0NM2aNePGjRs0atQIT09PhgwZQv/+/Z26loeHBzVq1GD9+vWOKfDR0dGsX7+efv36xalfrlw59uzZE6ts9OjRXL16lZkzZ6rbS+S/XLoEefJY3xcsaK0RVLKktWWGiEgm5HQi5OLiwqhRoxg6dCh///03165do0KFCmTPnj1JAQwePJigoCBq1qxJ7dq1mTFjBtevX6dbt24AdOnShSJFijBp0iS8vLyoWLFirPNz5swJEKdcRO7y2WfQvTvMmQOBgVZZuXL2xiQiYrMkryzt4eFBhQoV7juAwMBALly4wNixYzl79ixVq1ZlzZo1jgHUx48fx9U1Taz7KJI+hYVZ0+Hff986XrgQ2rWz1ggSEcnkXIwxxpkTmjRpgss9/gP94Ycf7juolBQaGoqvry8hISHkyJHD7nAkib76Cp580hre8uuvdkeThv31l5X07NplHQ8fDq++CncsfSEikh6k1O9vp1uEqlatGuv41q1b7Nq1i7179xIUFJRccYnI/fr0U+jVC65dg7x54aOPoEULu6MSEUlTnE6Epk+fHm/5+PHjuXbt2n0HJCLJYPfu23uDNWoEn3wCRYrYG5OISBqUbINvnnvuuXtuiyEiqahyZWuz1DFjYP16JUEiIglI8mDpu23ZsgUvL6/kupyIOGvxYmjYEIoVs47feksDokVE/oPTidAzzzwT69gYw5kzZ/j999+dXlBRJLHCw+HUKetx8qTVyCH/7/p16N8fFiyAevWstYHc3ZUEiYgkgtOJkK+vb6xjV1dXypYty6uvvsqjjz6abIFJ5nHt2u0E5+5HTPmFC/Gfm+kn/u3bZ80K278fXF2heXPrq4iIJIpTiVBUVBTdunWjUqVK5MqVK6VikgzCGLhyJeHkJuYREpK463l5QdGisR9du6bkO0jDjLFagPr1g5s3oVAha0B048Z2RyYikq44lQi5ubnx6KOPcuDAASVCmVx0tNVKc68E5+RJ63d0YuTIcTu5KVIkbsJTpAjkzq3eHsDqCuvdGz7+2Dpu3hwWLYL8+e2NS0QkHXK6a6xixYocOXKEEiVKpEQ8ksaEhVnLzxw8GDvBOX0abt1K3DXy5Imb2NyZ8BQpoi4up7i6WtPj3dzg9dfhlVfUHSYikkROJ0Kvv/46Q4YM4bXXXqNGjRpky5Yt1vNarTlj+fhja02++Li4WPt23qslp3Bh8PZO3ZgzJGOsh6ur9YEuW2Y1yTVoYHdkIiLpWqIToVdffZWXX36Zxx57DIAnn3wy1lYbxhhcXFyIiopK/ijFNv/+a32tWBGefz52wlOokHZqSBUhIVY2WqkSjB5tlZUtaz1EROS+JDoRmjBhAr179+bHH39MyXgkjapeHQYNsjuKTGj7dmun+MOHYdUqa/f4QoXsjkpEJMNIdCIUszfrww8/nGLBiMj/Mwbee89aHToiAooXhyVLlASJiCQzp8YI3WvXeRFJJleuWC0/K1ZYx61bw/z5oJmaIiLJzqlEqEyZMv+ZDF2+fPm+AhLJ1CIjrdWhDxywBmC9/ba1arT+CBERSRFOJUITJkyIs7K0iCSjLFlg4EBrn7ClS6FmTbsjEhHJ0JxKhNq3b09+LdomkrwuX4YzZ+DBB63jXr3guefgrqUpREQk+SV6FTaNDxJJAb/8AlWrwhNPWGODwOoGUxIkIpIqEp0IxcwaE5FkEB0NkydDo0Zw4oQ1Huj8ebujEhHJdBLdNRYdHZ2ScYhkHhcuQFAQfPutddyhA3zwAfj42BuXiEgm5PQWGyJyH37+2Up8Tp8GLy94911rqry6nkVEbKFESCQ1TZtmJUHlyln7hVWqZHdEIiKZmhIhkdQ0bx6ULAmvvgrZs9sdjYhIppfowdIikgQ//AAvv2xtmQGQJ4/VKqQkSEQkTVCLkEhKiIqyWn1ee81KgurUgXbt7I5KRETuokRIJLmdPg2dOsGGDdZx9+7WOkEiIpLmKBESSU7ffWetCn3hgrUo4gcfWEmRiIikSRojJJJcpkyBFi2sJKhKFdixQ0mQiEgap0RIJLlUq2Z9ffFF+PVXKFPG3nhEROQ/qWtM5H6cPw8xGxEHBMCePbc3TxURkTRPLUIiSXHrFgwdarX6HD58u1xJkIhIuqJESMRZ//wDDRvC229DSAh89ZXdEYmISBKpa0zEGV98Ad26wZUr4OsL8+fDM8/YHZWIiCSRWoREEiMiAl56CZ5+2kqCateGnTuVBImIpHNKhEQS4733YOZM6/vBg2HjRihRwt6YRETkvqlrTCQx+vWD77+HPn2gVSu7oxERkWSiFiGR+ISFWZuj3rplHXt4wLffKgkSEclg1CIkcre//oLAQGsM0IULMGmS3RGJiEgKUYuQyJ2WLIHq1a0kKG9eaNTI7ohERCQFqUUokwsPhxMn4Phxa3mc+L5mCjdvWrPC5syxjhs2hE8/hSJFbA1LRERSlhKhDMwYa6Z3QknOP//A2bP/fR0PD2sv0Qzr0CF49llrewwXFxg1CsaNgyz68RARyej0P306FhUFZ84k3JLzzz9w9ep/X8fbG4oVg+LF4/9atCi4u6f8+7FNdDQcOWLtGbZ4sbVnmIiIZApKhNKwGzesbquY1pu7k5yTJyEy8r+vkzdv/ElOzPd581oNIZlKdDS4/v8QuXLlYMUKqFQJChWyNy4REUlVSoTSiJMn4Z13rP07Y5KdCxf++7wsWawWm4Rac4oVg6xZUz7+dGXfPujQwVokMWYw9KOP2huTiIjYQolQGvH227cXLr6Tj8+9W3MKFQI3t9SPN10yxtobrH9/a3D0yy/D1q2ZsDlMRERiKBFKI65ds74+/bS1p2dMwuPrq9/TyeLqVXjxRWsMEFgtQB99pA9XRCSTUyKUxtSqpcWLk90ff0C7dtbsMDc3eO01GDbs9hghERHJtJQIScZ24ADUqWMtmFSkiLVgYoMGdkclIiJphBIhydjKlYMnn4Tr12HhQmuKnIiIyP9TIiQZz86dUKIE5MxpjQFauBA8PdUVJiIiceg3g2QcxlhT4h96CHr0sI7BWjFSSZCIiMRDLUKSMVy5At27WwsjgrXSZFiYlQSJiIgkQH8mS/q3dStUq2YlQe7uMGMGrFypJEhERP6TEiFJv4yB6dOtWWDHjlnjgjZvhoEDtT6QiIgkihIhSb9CQmDaNLh1C9q0gR07rIWYREREEkljhCT9ypkTPv3UWjCxTx+1AomIiNOUCEn6ER1tbcpWsCB06WKVNWigBRJFRCTJlAhJ+nDhAgQFwbffQtas0KQJ+PnZHZWIiKRzSoQk7du4Edq3h9OnwcvLmhVWtKjdUYmISAagwdKSdkVHwxtvQOPGVhJUtiz89hv07KnxQCIikizUIiRpU1QUPP44rF1rHXfuDO+/D9mz2xuXiIhkKGoRkrTJzQ1q1rTGAy1YAIsWKQkSEZFkp0RI0o6oKGtQdIzx42HXLuja1aaAREQko0sTidCsWbPw9/fHy8uLOnXqsHXr1gTrzp07l4YNG5IrVy5y5cpFQEDAPetLOnHmDDzyCLRsCeHhVlmWLFC6tL1xiYhIhmZ7IrR06VIGDx7MuHHj2LFjB1WqVKF58+acP38+3vobNmygQ4cO/Pjjj2zZsgU/Pz8effRRTp06lcqRS7L57juoUgV+/BH+/NNaIFFERCQV2J4ITZs2jZ49e9KtWzcqVKjA7NmzyZo1K/Pnz4+3/uLFi+nTpw9Vq1alXLlyfPjhh0RHR7N+/fpUjlzuW2QkjBoFLVpYXWKVK8P27VC7tt2RiYhIJmFrIhQREcH27dsJCAhwlLm6uhIQEMCWLVsSdY0bN25w69YtcufOHe/z4eHhhIaGxnpIGnDyJDRtChMnWpunvvAC/PqrNUVeREQkldiaCF28eJGoqCgKFCgQq7xAgQKcPXs2UdcYNmwYhQsXjpVM3WnSpEn4+vo6Hn5ajTht6NnTWijRxweWLIHZs8Hb2+6oREQkk7G9a+x+vPnmmyxZsoSVK1fi5eUVb50RI0YQEhLieJw4cSKVo5R4zZplbZOxYwcEBtodjYiIZFK2LqiYN29e3NzcOHfuXKzyc+fOUbBgwXue+/bbb/Pmm2+ybt06KleunGA9T09PPD09kyVeuQ/Hj1uDonv0sI5LloQffrA3JhERyfRsbRHy8PCgRo0asQY6xwx8rlu3boLnvfXWW7z22musWbOGmjVrpkaocj9WrYKqVaFXLysZEhERSSNs32Jj8ODBBAUFUbNmTWrXrs2MGTO4fv063bp1A6BLly4UKVKESZMmATB58mTGjh3LJ598gr+/v2MsUfbs2cmulYfTlogIGDbM2iQVoFYtrQskIiJpiu2JUGBgIBcuXGDs2LGcPXuWqlWrsmbNGscA6uPHj+Pqervh6n//+x8RERE8++yzsa4zbtw4xo8fn5qhy70cPWqN/dm2zToeNAjefBM8POyNS0RE5A62J0IA/fr1o1+/fvE+t2HDhljHx44dS/mA5P588YW1LUZICOTKBcHB8OSTNgclIiISV5pIhCSDCQ21kqC6da2p8cWK2R2RiIhIvJQISfKIirJ2jAfo0gW8vODpp8Hd3d64RERE7iFdryMkacSSJVCpEly8eLusXTslQSIikuYpEZKku3nT2hqjQwc4cACmTbM7IhEREaeoa0yS5s8/rVafPXvAxQVGjgTN2hMRkXRGiZA476OP4MUX4fp1yJ8fPv4YHnnE7qhEREScpkRInPPBB9C7t/V9kyaweDEUKmRvTCIiIkmkMULinPbt4YEHrG6w779XEiQiIumaWoTk3oyxNkdt2tQaC+TrC7t3g7e33ZGJiIjcN7UIScKuXYOgIAgIgNmzb5crCRIRkQxCLUISv927rVlhBw+Cq6s1MFpERCSDUSIksRkDc+bAwIEQHg5FisCnn0LDhnZHJiIikuyUCMltoaHQqxcsXWodt2wJixZB3rz2xiUiIpJCNEZIbtu7F5Yvt/YMe+st+PprJUEiIpKhqUVIbqtXD957D6pWtXaOFxERyeDUIpSZXbkCnTtb+4TFePFFJUEiIpJpqEUos9q2DQID4ehR2L8ffv/dWidIREQkE1GLUGZjDMyYAfXrW0mQv7+1RpCSIBERyYTUIpSZXL4M3brBqlXW8TPPwLx5kDOnrWGJiIjYRYlQZnH0KDRuDMePg4cHTJsGffqoJUhERDI1JUKZhZ8fFCsG7u6wbBlUr253RCIiIrZTIpSRXboEPj5WC1CWLNYaQVmzQo4cdkcmIiKSJmiwdEa1cSNUqQLDht0uK1hQSZCIiMgdlAhlNNHRMHEiNGkCp07BmjXaMFVERCQBSoQykvPnoUULGDUKoqLguees9YKyZbM7MhERkTRJY4Qyih9/hI4d4exZ8PaGWbOga1fNChMREbkHJUIZQWgotGkD//4LFSpYs8IefNDuqERERNI8JUIZQY4c8MEH8O238O676goTERFJJCVC6dW6deDqCk2bWsdt21oPERERSTQNlk5vIiNh9Gh49FHo0AHOnLE7IhERkXRLLULpyalTVvKzcaN13Lq19gkTERG5D0qE0otvv4UuXeDiRcieHebOhfbt7Y5KREQkXVPXWFoXHW2tDv3YY1YSVK0a7NihJEhERCQZKBFK61xdrbWBAPr2hV9+gdKl7Y1JREQkg1DXWFoVGWltlArW4oht28ITT9gbk4hICjPGEBkZSVRUlN2hiA3c3d1xc3NL1ddUIpSKoqOt8c6HD8ORI9bXmO/37bPquEZGwODh8Pff8OWX1srQ2bMrCRKRDC8iIoIzZ85w48YNu0MRm7i4uFC0aFGyZ8+eaq+pRCiZ3bhhJTYxic6dX48ehYiIhM8t63GUFxcHwsFtVsGGDdbmqSIiGVx0dDRHjx7Fzc2NwoUL4+HhgYu2CMpUjDFcuHCBkydPUrp06VRrGVIi5CRj4Ny5+BOdw4dvD+dJSJYs4O8PJUtCqVK3v1Y7uoLirz6Py8EQa0p8cLCSIBHJNCIiIoiOjsbPz4+sWbPaHY7YJF++fBw7doxbt24pEbJTeDj880/8ic6RI1arz734+lrJzZ2JTszXokVvD/1xvNiQIfDee9bxQw/BkiVQvHiKvT8RkbTK1VVzeDIzO1oBM20idPkyHDoU/3idEyeslp+EuLqCn5+V3Nyd6JQsCblzOxFIp07w+efW90OHwhtvgLv7fb03ERERSZxMmwiVKHHv57Nmjb9Fp2RJq2vLwyOZAhk2zJoSP3cuPP54Ml1UREREEiNTt0EWKgT161sLNo8fDx99BJs3W+N8rl2D3bvhiy9g6lRrCZ8WLaBMmftMgm7ehJ9+un1cq5bVDKUkSEQk3dqyZQtubm48Hs//5Rs2bMDFxYUrV67Eec7f358ZM2bEKvvxxx957LHHyJMnD1mzZqVChQq8/PLLnDp1KoWih7CwMPr27UuePHnInj07bdq04dy5c/c8x8XFJd7HlClTgNvvO77Htm3bUuy9OCvTJkILFsDp07BpEyxcCOPGwXPPQb16UKCANWs92R08aI0Bat4cdu26Xe7llQIvJiIiqWXevHn079+fn3/+mdOnTyf5Oh988AEBAQEULFiQzz//nP379zN79mxCQkKYOnVqMkYc26BBg/jqq69Yvnw5P/30E6dPn+aZZ5655zlnzpyJ9Zg/fz4uLi60adMGgHr16sWp06NHD0qUKEHNmjVT7L04K9N2jWVJ7Xe+eDG88AJcvw758kE8fxmIiEj6c+3aNZYuXcrvv//O2bNnCQ4OZuTIkU5f5+TJkwwYMIABAwYwffp0R7m/vz+NGjWKt0UpOYSEhDBv3jw++eQTmjZtCsCCBQsoX748v/76Kw899FC85xUsWDDW8ZdffkmTJk0oWbIkAB4eHrHq3Lp1iy+//JL+/funqaURMm2LUKq5cQN69LCam65fh8aNrdagxo1tDkxEJG0zxvpvM7Uf95osE59ly5ZRrlw5ypYty3PPPcf8+fMxzl4EWL58OREREbzyyivxPp8zZ84Ez23ZsiXZs2dP8PHggw8meO727du5desWAQEBjrJy5cpRrFgxtmzZkqjYz507x+rVq+nevXuCdVatWsWlS5fo1q1boq6ZWjJti1Cq2L8f2rWzlo12cYGxY2HMGEjl5cNFRNKjGzeshfVT27VrkC1b4uvPmzeP5557DoAWLVoQEhLCTz/9RGMn/+D966+/yJEjB4UKFXLqPIAPP/yQmzdvJvi8+z1mI589exYPD484iVaBAgU4+1+L4/2/hQsX4uPjc8/utHnz5tG8eXOKFi2aqGumFiVCKenLL60kqGBBq2vs/5scRUQkYzh48CBbt25l5cqVAGTJkoXAwEDmzZvndCJkjElyl1GRIkWSdF5ymT9/Pp06dcIrgTGvJ0+eZO3atSxbtiyVI/tvSoRS0iuvWO2s/ftbI7BFRCTRsma1WmfseN3EmjdvHpGRkRQuXNhRZozB09OT9957D19fX3LkyAFYY3HubnW5cuUKvr6+AJQpU4aQkBDOnDnjdKtQy5Yt2bhxY4LPFy9enH0xm1repWDBgkRERHDlypVY8Z07dy7OOKD4bNy4kYMHD7J06dIE6yxYsIA8efLw5JNP/uf1UpsSoeS0Zw+8+iosWgTe3lYX2Ouv2x2ViEi65OLiXBdVaouMjGTRokVMnTqVRx99NNZzrVu35tNPP6V3796ULl0aV1dXtm/fTvE7dg04cuQIISEhlClTBoBnn32W4cOH89Zbb8UaLB3j7kTlTvfTNVajRg3c3d1Zv369Y8bXwYMHOX78OHXr1k3wvBjz5s2jRo0aVKlSJd7njTEsWLCALl263DMOuygRSg7GwIcfwoABEBZmrbo4ebLdUYmISAr6+uuv+ffff+nevbujVSdGmzZtmDdvHr1798bHx4cePXrw8ssvkyVLFipVqsSJEycYNmwYDz30EPXq1QPAz8+P6dOn069fP0JDQ+nSpQv+/v6cPHmSRYsWkT179gSn0N9P15ivry/du3dn8ODB5M6dmxw5ctC/f3/q1q0ba8ZYuXLlmDRpEk8//bSjLDQ0lOXLl99zav8PP/zA0aNH6dGjR5JjTEmaNXa/QkOhY0fo1ctKglq0sPYOExGRDG3evHkEBATESYLASoR+//13du/eDcDMmTMJCgpi2LBhPPjgg3Tt2pXKlSvz1VdfxRoX1KdPH7777jtOnTrF008/Tbly5ejRowc5cuRgSAr+bpk+fTpPPPEEbdq0oVGjRhQsWJAVK1bEqnPw4EFCQkJilS1ZsgRjDB06dEjw2vPmzaNevXqUK1cuRWK/Xy4mKXP80rHQ0FB8fX356KMQnnsux/1dbOdOa1bY339b3WATJ1pJkDYNFBFxSlhYGEePHqVEiRIJDriVjO9e/w5ifn+HhIQ4xl0lB3WNJdXKldC+PUREWDuwLlliLUstIiIi6YYSoaSqWdNa4KJ+fWu/jjx57I5IREREnKREyBmnTkHMgDQ/P9i61RoYnYaWChcREZHE02CWxDAGZs60kp5Vq26XlyqlJEhERCQdUyL0Xy5fhqefhpdessYD3ZkIiYiISLqmROhefv0VqlWztsrw8IB334W5c+2OSkQkw8pkE5nlLnbcfyVC8YmOhrffhoYN4fhxqwvsl1+gXz91hYmIpICYFYdv3LhhcyRip4iICADcUnFzcg2Wjs/PP8PQodb37dpZrUDJuGaBiIjE5ubmRs6cOTl//jwAWbNmTfIGpJI+RUdHc+HCBbJmzUqWLKmXnigRik/jxjBwIJQrBy+8oFYgEZFUELPBZ0wyJJmPq6srxYoVS9UkWIkQWF1hM2dChw4Qs9PujBm2hiQiktm4uLhQqFAh8ufPz61bt+wOR2zg4eGBayrvzpAmEqFZs2YxZcoUzp49S5UqVXj33XepXbt2gvWXL1/OmDFjOHbsGKVLl2by5Mk89thjSXvx8+ehc2f47jv4+mv4/nttkSEiYiM3N7dUHSMimZvtv/GXLl3K4MGDGTduHDt27KBKlSo0b948wabRX375hQ4dOtC9e3d27txJ69atad26NXv37nX+xTdsgKpVrSTI2xs6dVI3mIiISCZi+6arderUoVatWrz33nuANVjKz8+P/v37M3z48Dj1AwMDuX79Ol9//bWj7KGHHqJq1arMnj37P18vZtO2354ZQe0vJlvdYuXLw7JlULFi8r0xERERSTYptemqrS1CERERbN++nYCAAEeZq6srAQEBbNmyJd5ztmzZEqs+QPPmzROsn5ByKyZZSVC3brBtm5IgERGRTMjWMUIXL14kKiqKAgUKxCovUKAAf/75Z7znnD17Nt76Z8+ejbd+eHg44eHhjuOQkBAALrt7wXszrR3ko6IgNPR+3oqIiIikoND//z2d3B1ZaWKwdEqaNGkSEyZMiFNe4laYNTX+hRdsiEpERESS4tKlS/j6+ibb9WxNhPLmzYubmxvnzp2LVX7u3DnHehJ3K1iwoFP1R4wYweDBgx3HV65coXjx4hw/fjxZP0hxXmhoKH5+fpw4cSJZ+3slaXQ/0g7di7RD9yLtCAkJoVixYuTOnTtZr2trIuTh4UGNGjVYv349rVu3BqzB0uvXr6dfv37xnlO3bl3Wr1/PSy+95Cj7/vvvqVu3brz1PT098fT0jFPu6+urf9RpRI4cOXQv0hDdj7RD9yLt0L1IO5J7nSHbu8YGDx5MUFAQNWvWpHbt2syYMYPr16/TrVs3ALp06UKRIkWYNGkSAAMHDuThhx9m6tSpPP744yxZsoTff/+dOXPm2Pk2REREJB2yPREKDAzkwoULjB07lrNnz1K1alXWrFnjGBB9/PjxWNlfvXr1+OSTTxg9ejQjR46kdOnSfPHFF1TUrC8RERFxku2JEEC/fv0S7ArbsGFDnLK2bdvStm3bJL2Wp6cn48aNi7e7TFKX7kXaovuRduhepB26F2lHSt0L2xdUFBEREbGL7VtsiIiIiNhFiZCIiIhkWkqEREREJNNSIiQiIiKZVoZMhGbNmoW/vz9eXl7UqVOHrVu33rP+8uXLKVeuHF5eXlSqVIlvvvkmlSLN+Jy5F3PnzqVhw4bkypWLXLlyERAQ8J/3Tpzj7M9GjCVLluDi4uJY+FTun7P34sqVK/Tt25dChQrh6elJmTJl9H9VMnH2XsyYMYOyZcvi7e2Nn58fgwYNIiwsLJWizbh+/vlnWrVqReHChXFxceGLL774z3M2bNhA9erV8fT05IEHHiA4ONj5FzYZzJIlS4yHh4eZP3++2bdvn+nZs6fJmTOnOXfuXLz1N2/ebNzc3Mxbb71l9u/fb0aPHm3c3d3Nnj17UjnyjMfZe9GxY0cza9Yss3PnTnPgwAHTtWtX4+vra06ePJnKkWdMzt6PGEePHjVFihQxDRs2NE899VTqBJvBOXsvwsPDTc2aNc1jjz1mNm3aZI4ePWo2bNhgdu3alcqRZzzO3ovFixcbT09Ps3jxYnP06FGzdu1aU6hQITNo0KBUjjzj+eabb8yoUaPMihUrDGBWrlx5z/pHjhwxWbNmNYMHDzb79+837777rnFzczNr1qxx6nUzXCJUu3Zt07dvX8dxVFSUKVy4sJk0aVK89du1a2cef/zxWGV16tQxL7zwQorGmRk4ey/uFhkZaXx8fMzChQtTKsRMJSn3IzIy0tSrV898+OGHJigoSIlQMnH2Xvzvf/8zJUuWNBEREakVYqbh7L3o27evadq0aayywYMHm/r166donJlNYhKhV155xTz44IOxygIDA03z5s2deq0M1TUWERHB9u3bCQgIcJS5uroSEBDAli1b4j1ny5YtseoDNG/ePMH6kjhJuRd3u3HjBrdu3Ur2DfYyo6Tej1dffZX8+fPTvXv31AgzU0jKvVi1ahV169alb9++FChQgIoVKzJx4kSioqJSK+wMKSn3ol69emzfvt3RfXbkyBG++eYbHnvssVSJWW5Lrt/faWJl6eRy8eJFoqKiHNtzxChQoAB//vlnvOecPXs23vpnz55NsTgzg6Tci7sNGzaMwoULx/mHLs5Lyv3YtGkT8+bNY9euXakQYeaRlHtx5MgRfvjhBzp16sQ333zD33//TZ8+fbh16xbjxo1LjbAzpKTci44dO3Lx4kUaNGiAMYbIyEh69+7NyJEjUyNkuUNCv79DQ0O5efMm3t7eibpOhmoRkozjzTffZMmSJaxcuRIvLy+7w8l0rl69SufOnZk7dy558+a1O5xMLzo6mvz58zNnzhxq1KhBYGAgo0aNYvbs2XaHluls2LCBiRMn8v7777Njxw5WrFjB6tWree211+wOTZIoQ7UI5c2bFzc3N86dOxer/Ny5cxQsWDDecwoWLOhUfUmcpNyLGG+//TZvvvkm69ato3LlyikZZqbh7P04fPgwx44do1WrVo6y6OhoALJkycLBgwcpVapUygadQSXlZ6NQoUK4u7vj5ubmKCtfvjxnz54lIiICDw+PFI05o0rKvRgzZgydO3emR48eAFSqVInr16/Tq1cvRo0aFWuTcElZCf3+zpEjR6JbgyCDtQh5eHhQo0YN1q9f7yiLjo5m/fr11K1bN95z6tatG6s+wPfff59gfUmcpNwLgLfeeovXXnuNNWvWULNmzdQINVNw9n6UK1eOPXv2sGvXLsfjySefpEmTJuzatQs/P7/UDD9DScrPRv369fn7778dySjAoUOHKFSokJKg+5CUe3Hjxo04yU5Mgmq0dWeqSrbf386N4077lixZYjw9PU1wcLDZv3+/6dWrl8mZM6c5e/asMcaYzp07m+HDhzvqb9682WTJksW8/fbb5sCBA2bcuHGaPp9MnL0Xb775pvHw8DCfffaZOXPmjONx9epVu95ChuLs/bibZo0lH2fvxfHjx42Pj4/p16+fOXjwoPn6669N/vz5zeuvv27XW8gwnL0X48aNMz4+PubTTz81R44cMd99950pVaqUadeunV1vIcO4evWq2blzp9m5c6cBzLRp08zOnTvNP//8Y4wxZvjw4aZz586O+jHT54cOHWoOHDhgZs2apenzMd59911TrFgx4+HhYWrXrm1+/fVXx3MPP/ywCQoKilV/2bJlpkyZMsbDw8M8+OCDZvXq1akcccblzL0oXry4AeI8xo0bl/qBZ1DO/mzcSYlQ8nL2Xvzyyy+mTp06xtPT05QsWdK88cYbJjIyMpWjzpicuRe3bt0y48ePN6VKlTJeXl7Gz8/P9OnTx/z777+pH3gG8+OPP8b7OyDm8w8KCjIPP/xwnHOqVq1qPDw8TMmSJc2CBQucfl0XY9SWJyIiIplThhojJCIiIuIMJUIiIiKSaSkREhERkUxLiZCIiIhkWkqEREREJNNSIiQiIiKZlhIhERERybSUCIlILMHBweTMmdPuMJLMxcWFL7744p51unbtSuvWrVMlHhFJ25QIiWRAXbt2xcXFJc7j77//tjs0goODHfG4urpStGhRunXrxvnz55Pl+mfOnKFly5YAHDt2DBcXF3bt2hWrzsyZMwkODk6W10vI+PHjHe/Tzc0NPz8/evXqxeXLl526jpI2kZSVoXafF5HbWrRowYIFC2KV5cuXz6ZoYsuRIwcHDx4kOjqaP/74g27dunH69GnWrl1739dOaNfwO/n6+t736yTGgw8+yLp164iKiuLAgQM8//zzhISEsHTp0lR5fRH5b2oREsmgPD09KViwYKyHm5sb06ZNo1KlSmTLlg0/Pz/69OnDtWvXErzOH3/8QZMmTfDx8SFHjhzUqFGD33//3fH8pk2baNiwId7e3vj5+TFgwACuX79+z9hcXFwoWLAghQsXpmXLlgwYMIB169Zx8+ZNoqOjefXVVylatCienp5UrVqVNWvWOM6NiIigX79+FCpUCC8vL4oXL86kSZNiXTuma6xEiRIAVKtWDRcXFxo3bgzEbmWZM2cOhQsXjrWzO8BTTz3F888/7zj+8ssvqV69Ol5eXpQsWZIJEyYQGRl5z/eZJUsWChYsSJEiRQgICKBt27Z8//33juejoqLo3r07JUqUwNvbm7JlyzJz5kzH8+PHj2fhwoV8+eWXjtalDRs2AHDixAnatWtHzpw5yZ07N0899RTHjh27ZzwiEpcSIZFMxtXVlXfeeYd9+/axcOFCfvjhB1555ZUE63fq1ImiRYuybds2tm/fzvDhw3F3dwfg8OHDtGjRgjZt2rB7926WLl3Kpk2b6Nevn1MxeXt7Ex0dTWRkJDNnzmTq1Km8/fbb7N69m+bNm/Pkk0/y119/AfDOO++watUqli1bxsGDB1m8eDH+/v7xXnfr1q0ArFu3jjNnzrBixYo4ddq2bculS5f48ccfHWWXL19mzZo1dOrUCYCNGzfSpUsXBg4cyP79+/nggw8IDg7mjTfeSPR7PHbsGGvXrsXDw8NRFh0dTdGiRVm+fDn79+9n7NixjBw5kmXLlgEwZMgQ2rVrR4sWLThz5gxnzpyhXr163Lp1i+bNm+Pj48PGjRvZvHkz2bNnp0WLFkRERCQ6JhGBDLn7vEhmFxQUZNzc3Ey2bNkcj2effTbeusuXLzd58uRxHC9YsMD4+vo6jn18fExwcHC853bv3t306tUrVtnGjRuNq6uruXnzZrzn3H39Q4cOmTJlypiaNWsaY4wpXLiweeONN2KdU6tWLdOnTx9jjDH9+/c3TZs2NdHR0fFeHzArV640xhhz9OhRA5idO3fGqhMUFGSeeuopx/FTTz1lnn/+ecfxBx98YAoXLmyioqKMMcY0a9bMTJw4MdY1PvroI1OoUKF4YzDGmHHjxhlXV1eTLVs24+Xl5dhJe9q0aQmeY4wxffv2NW3atEkw1pjXLlu2bKzPIDw83Hh7e5u1a9fe8/oiEpvGCIlkUE2aNOF///uf4zhbtmyA1ToyadIk/vzzT0JDQ4mMjCQsLIwbN26QNWvWONcZPHgwPXr04KOPPnJ075QqVQqwus12797N4sWLHfWNMURHR3P06FHKly8fb2whISFkz56d6OhowsLCaNCgAR9++CGhoaGcPn2a+vXrx6pfv359/vjjD8Dq1nrkkUcoW7YsLVq04IknnuDRRx+9r8+qU6dO9OzZk/fffx9PT08WL15M+/btcXV1dbzPzZs3x2oBioqKuufnBlC2bFlWrVpFWFgYH3/8Mbt27aJ///6x6syaNYv58+dz/Phxbt68SUREBFWrVr1nvH/88Qd///03Pj4+scrDwsI4fPhwEj4BkcxLiZBIBpUtWzYeeOCBWGXHjh3jiSee4MUXX+SNN94gd+7cbNq0ie7duxMRERHvL/Tx48fTsWNHVq9ezbfffsu4ceNYsmQJTz/9NNeuXeOFF15gwIABcc4rVqxYgrH5+PiwY8cOXF1dKVSoEN7e3gCEhob+5/uqXr06R48e5dtvv2XdunW0a9eOgIAAPvvss/88NyGtWrXCGMPq1aupVasWGzduZPr06Y7nr127xoQJE3jmmWfinOvl5ZXgdT08PBz34M033+Txxx9nwoQJvPbaawAsWbKEIUOGMHXqVOrWrYuPjw9Tpkzht99+u2e8165do0aNGrES0BhpZUC8SHqhREgkE9m+fTvR0dFMnTrV0doRMx7lXsqUKUOZMmUYNGgQHTp0YMGCBTz99NNUr16d/fv3x0m4/ourq2u85+TIkYPChQuzefNmHn74YUf55s2bqV27dqx6gYGBBAYG8uyzz9KiRQsuX75M7ty5Y10vZjxOVFTUPePx8vLimWeeYfHixfz999+ULVuW6tWrO56vXr06Bw8edPp93m306NE0bdqUF1980fE+69WrR58+fRx17m7R8fDwiBN/9erVWbp0Kfnz5ydHjhz3FZNIZqfB0iKZyAMPPMCtW7d49913OXLkCB999BGzZ89OsP7Nmzfp168fGzZs4J9//mHz5s1s27bN0eU1bNgwfvnlF/r168euXbv466+/+PLLL50eLH2noUOHMnnyZJYuXcrBgwcZPnw4u3btYuDAgQBMmzaNTz/9lD///JNDhw6xfPlyChYsGO8ikPnz58fb25s1a9Zw7tw5QkJCEnzdTp06sXr1aubPn+8YJB1j7NixLFq0iAkTJrBv3z4OHDjAkiVLGD16tFPvrW7dulSuXJmJEycCULp0aX7//XfWrl3LoUOHGDNmDNu2bYt1jr+/P7t37+bgwYNcvHiRW7du0alTJ/LmzctTTz3Fxo0bOXr0KBs2bGDAgAGcPHnSqZhEMj27BymJSPKLb4BtjGnTpplChQoZb29v07x5c7No0SIDmH///dcYE3swc3h4uGnfvr3x8/MzHh4epnDhwqZfv36xBkJv3brVPPLIIyZ79uwmW7ZspnLlynEGO9/p7sHSd4uKijLjx483RYoUMe7u7qZKlSrm22+/dTw/Z84cU7VqVZMtWzaTI0cO06xZM7Njxw7H89wxWNoYY+bOnWv8/PyMq6urefjhhxP8fKKiokyhQoUMYA4fPhwnrjVr1ph69eoZb29vkyNHDlO7dm0zZ86cBN/HuHHjTJUqVeKUf/rpp8bT09McP37chIWFma5duxpfX1+TM2dO8+KLL5rhw4fHOu/8+fOOzxcwP/74ozHGmDNnzpguXbqYvHnzGk9PT1OyZEnTs2dPExISkmBMIhKXizHG2JuKiYiIiNhDXWMiIiKSaSkREhERkUxLiZCIiIhkWkqEREREJNNSIiQiIiKZlhIhERERybSUCImIiEimpURIREREMi0lQiIiIpJpKRESERGRTEuJkIiIiGRaSoREREQk0/o/JQ2vFvyStMkAAAAASUVORK5CYII=\n"
          },
          "metadata": {}
        }
      ],
      "source": [
        "# Compute predicted probabilities on the test set\n",
        "probs = bert_predict(bert_classifier, val_dataloader)\n",
        "\n",
        "# Evaluate the Bert classifier\n",
        "evaluate_roc(probs, y_val)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 62,
      "metadata": {
        "id": "2PKo6SLH-E-2",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "738be740-8e51-4ef0-a99c-5d5882d53c3c"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Start training...\n",
            "\n",
            " Epoch  |  Batch  |  Train Loss  |  Val Loss  |  Val Acc  |  Elapsed \n",
            "----------------------------------------------------------------------\n",
            "   1    |   20    |   0.107408   |     -      |     -     |   6.90   \n",
            "   1    |   40    |   0.021288   |     -      |     -     |   6.78   \n",
            "   1    |   60    |   0.012286   |     -      |     -     |   6.80   \n",
            "   1    |   80    |   0.002431   |     -      |     -     |   6.81   \n",
            "   1    |   100   |   0.011815   |     -      |     -     |   6.85   \n",
            "   1    |   120   |   0.011677   |     -      |     -     |   6.82   \n",
            "   1    |   140   |   0.001763   |     -      |     -     |   6.85   \n",
            "   1    |   160   |   0.011791   |     -      |     -     |   6.86   \n",
            "   1    |   180   |   0.011796   |     -      |     -     |   6.90   \n",
            "   1    |   200   |   0.011880   |     -      |     -     |   6.89   \n",
            "   1    |   220   |   0.022405   |     -      |     -     |   6.91   \n",
            "   1    |   240   |   0.012045   |     -      |     -     |   6.91   \n",
            "   1    |   260   |   0.011755   |     -      |     -     |   6.95   \n",
            "   1    |   280   |   0.041655   |     -      |     -     |   6.96   \n",
            "   1    |   300   |   0.003501   |     -      |     -     |   6.96   \n",
            "   1    |   320   |   0.011803   |     -      |     -     |   6.98   \n",
            "   1    |   340   |   0.011604   |     -      |     -     |   6.99   \n",
            "   1    |   360   |   0.001391   |     -      |     -     |   7.01   \n",
            "   1    |   380   |   0.001065   |     -      |     -     |   7.00   \n",
            "   1    |   400   |   0.022989   |     -      |     -     |   7.01   \n",
            "   1    |   420   |   0.021269   |     -      |     -     |   7.01   \n",
            "   1    |   440   |   0.012812   |     -      |     -     |   7.02   \n",
            "   1    |   460   |   0.001496   |     -      |     -     |   7.02   \n",
            "   1    |   480   |   0.001115   |     -      |     -     |   7.05   \n",
            "   1    |   500   |   0.000905   |     -      |     -     |   7.04   \n",
            "   1    |   520   |   0.011994   |     -      |     -     |   7.03   \n",
            "   1    |   540   |   0.022897   |     -      |     -     |   7.05   \n",
            "   1    |   560   |   0.011734   |     -      |     -     |   7.00   \n",
            "   1    |   580   |   0.031106   |     -      |     -     |   7.05   \n",
            "   1    |   600   |   0.001529   |     -      |     -     |   7.03   \n",
            "   1    |   620   |   0.011958   |     -      |     -     |   7.05   \n",
            "   1    |   640   |   0.011906   |     -      |     -     |   7.05   \n",
            "   1    |   660   |   0.001282   |     -      |     -     |   7.06   \n",
            "   1    |   680   |   0.000941   |     -      |     -     |   7.05   \n",
            "   1    |   700   |   0.000788   |     -      |     -     |   7.05   \n",
            "   1    |   720   |   0.012223   |     -      |     -     |   7.07   \n",
            "   1    |   740   |   0.012048   |     -      |     -     |   7.03   \n",
            "   1    |   760   |   0.012101   |     -      |     -     |   7.07   \n",
            "   1    |   780   |   0.022134   |     -      |     -     |   7.05   \n",
            "   1    |   800   |   0.001645   |     -      |     -     |   7.10   \n",
            "   1    |   820   |   0.001071   |     -      |     -     |   7.14   \n",
            "   1    |   840   |   0.011213   |     -      |     -     |   7.25   \n",
            "   1    |   860   |   0.001071   |     -      |     -     |   7.19   \n",
            "   1    |   880   |   0.023016   |     -      |     -     |   7.18   \n",
            "   1    |   900   |   0.001408   |     -      |     -     |   7.15   \n",
            "   1    |   920   |   0.011096   |     -      |     -     |   7.13   \n",
            "   1    |   940   |   0.012167   |     -      |     -     |   7.11   \n",
            "   1    |   960   |   0.012245   |     -      |     -     |   7.09   \n",
            "   1    |   980   |   0.010323   |     -      |     -     |   7.08   \n",
            "   1    |  1000   |   0.001087   |     -      |     -     |   7.10   \n",
            "   1    |  1020   |   0.000786   |     -      |     -     |   7.08   \n",
            "   1    |  1040   |   0.012084   |     -      |     -     |   7.12   \n",
            "   1    |  1060   |   0.000840   |     -      |     -     |   7.10   \n",
            "   1    |  1080   |   0.012075   |     -      |     -     |   7.11   \n",
            "   1    |  1100   |   0.000820   |     -      |     -     |   7.13   \n",
            "   1    |  1120   |   0.000729   |     -      |     -     |   7.12   \n",
            "   1    |  1140   |   0.012214   |     -      |     -     |   7.12   \n",
            "   1    |  1160   |   0.023427   |     -      |     -     |   7.12   \n",
            "   1    |  1180   |   0.001365   |     -      |     -     |   7.15   \n",
            "   1    |  1200   |   0.011730   |     -      |     -     |   7.14   \n",
            "   1    |  1220   |   0.022998   |     -      |     -     |   7.12   \n",
            "   1    |  1240   |   0.001956   |     -      |     -     |   7.19   \n",
            "   1    |  1260   |   0.011924   |     -      |     -     |   7.18   \n",
            "   1    |  1280   |   0.001022   |     -      |     -     |   7.14   \n",
            "   1    |  1300   |   0.032295   |     -      |     -     |   7.12   \n",
            "   1    |  1320   |   0.013046   |     -      |     -     |   7.14   \n",
            "   1    |  1340   |   0.020138   |     -      |     -     |   7.10   \n",
            "   1    |  1360   |   0.010670   |     -      |     -     |   7.13   \n",
            "   1    |  1380   |   0.012222   |     -      |     -     |   7.08   \n",
            "   1    |  1400   |   0.011919   |     -      |     -     |   7.11   \n",
            "   1    |  1420   |   0.011905   |     -      |     -     |   7.13   \n",
            "   1    |  1440   |   0.001212   |     -      |     -     |   7.09   \n",
            "   1    |  1460   |   0.022726   |     -      |     -     |   7.10   \n",
            "   1    |  1480   |   0.011091   |     -      |     -     |   7.10   \n",
            "   1    |  1500   |   0.001389   |     -      |     -     |   7.09   \n",
            "   1    |  1520   |   0.011803   |     -      |     -     |   7.07   \n",
            "   1    |  1540   |   0.022479   |     -      |     -     |   7.11   \n",
            "   1    |  1560   |   0.039983   |     -      |     -     |   7.08   \n",
            "   1    |  1580   |   0.003594   |     -      |     -     |   7.10   \n",
            "   1    |  1600   |   0.011964   |     -      |     -     |   7.11   \n",
            "   1    |  1620   |   0.001126   |     -      |     -     |   7.10   \n",
            "   1    |  1640   |   0.022425   |     -      |     -     |   7.18   \n",
            "   1    |  1660   |   0.001498   |     -      |     -     |   7.19   \n",
            "   1    |  1680   |   0.011647   |     -      |     -     |   7.21   \n",
            "   1    |  1700   |   0.012079   |     -      |     -     |   7.23   \n",
            "   1    |  1720   |   0.001235   |     -      |     -     |   7.22   \n",
            "   1    |  1740   |   0.011502   |     -      |     -     |   7.18   \n",
            "   1    |  1760   |   0.021528   |     -      |     -     |   7.19   \n",
            "   1    |  1780   |   0.029810   |     -      |     -     |   7.15   \n",
            "   1    |  1800   |   0.001408   |     -      |     -     |   7.14   \n",
            "   1    |  1820   |   0.011702   |     -      |     -     |   7.20   \n",
            "   1    |  1840   |   0.022215   |     -      |     -     |   7.17   \n",
            "   1    |  1860   |   0.001661   |     -      |     -     |   7.24   \n",
            "   1    |  1880   |   0.001176   |     -      |     -     |   7.24   \n",
            "   1    |  1900   |   0.012070   |     -      |     -     |   7.24   \n",
            "   1    |  1920   |   0.011554   |     -      |     -     |   7.25   \n",
            "   1    |  1940   |   0.012044   |     -      |     -     |   7.22   \n",
            "   1    |  1960   |   0.001337   |     -      |     -     |   7.25   \n",
            "   1    |  1980   |   0.000981   |     -      |     -     |   7.21   \n",
            "   1    |  2000   |   0.011899   |     -      |     -     |   7.19   \n",
            "   1    |  2020   |   0.000903   |     -      |     -     |   7.21   \n",
            "   1    |  2040   |   0.000779   |     -      |     -     |   7.20   \n",
            "   1    |  2060   |   0.000714   |     -      |     -     |   7.22   \n",
            "   1    |  2080   |   0.011852   |     -      |     -     |   7.21   \n",
            "   1    |  2100   |   0.000870   |     -      |     -     |   7.21   \n",
            "   1    |  2120   |   0.010819   |     -      |     -     |   7.18   \n",
            "   1    |  2140   |   0.000792   |     -      |     -     |   7.18   \n",
            "   1    |  2160   |   0.011145   |     -      |     -     |   7.20   \n",
            "   1    |  2180   |   0.001082   |     -      |     -     |   7.19   \n",
            "   1    |  2200   |   0.000609   |     -      |     -     |   7.21   \n",
            "   1    |  2220   |   0.012627   |     -      |     -     |   7.17   \n",
            "   1    |  2240   |   0.012277   |     -      |     -     |   7.23   \n",
            "   1    |  2260   |   0.000681   |     -      |     -     |   7.14   \n",
            "   1    |  2280   |   0.000562   |     -      |     -     |   7.18   \n",
            "   1    |  2300   |   0.034424   |     -      |     -     |   7.19   \n",
            "   1    |  2320   |   0.000975   |     -      |     -     |   7.22   \n",
            "   1    |  2340   |   0.000848   |     -      |     -     |   7.20   \n",
            "   1    |  2360   |   0.012270   |     -      |     -     |   7.20   \n",
            "   1    |  2380   |   0.012117   |     -      |     -     |   7.22   \n",
            "   1    |  2400   |   0.009234   |     -      |     -     |   7.20   \n",
            "   1    |  2420   |   0.010326   |     -      |     -     |   7.19   \n",
            "   1    |  2440   |   0.034909   |     -      |     -     |   7.22   \n",
            "   1    |  2460   |   0.012183   |     -      |     -     |   7.25   \n",
            "   1    |  2480   |   0.001510   |     -      |     -     |   7.24   \n",
            "   1    |  2500   |   0.010709   |     -      |     -     |   7.22   \n",
            "   1    |  2520   |   0.022701   |     -      |     -     |   7.24   \n",
            "   1    |  2540   |   0.009455   |     -      |     -     |   7.23   \n",
            "   1    |  2560   |   0.001348   |     -      |     -     |   7.22   \n",
            "   1    |  2580   |   0.000883   |     -      |     -     |   7.23   \n",
            "   1    |  2600   |   0.000721   |     -      |     -     |   7.22   \n",
            "   1    |  2620   |   0.000643   |     -      |     -     |   7.21   \n",
            "   1    |  2640   |   0.000560   |     -      |     -     |   7.24   \n",
            "   1    |  2660   |   0.010121   |     -      |     -     |   7.24   \n",
            "   1    |  2680   |   0.000737   |     -      |     -     |   7.23   \n",
            "   1    |  2700   |   0.014287   |     -      |     -     |   7.23   \n",
            "   1    |  2720   |   0.002365   |     -      |     -     |   7.25   \n",
            "   1    |  2740   |   0.020444   |     -      |     -     |   7.24   \n",
            "   1    |  2760   |   0.012758   |     -      |     -     |   7.22   \n",
            "   1    |  2780   |   0.012191   |     -      |     -     |   7.21   \n",
            "   1    |  2800   |   0.022802   |     -      |     -     |   7.22   \n",
            "   1    |  2820   |   0.012022   |     -      |     -     |   7.19   \n",
            "   1    |  2840   |   0.011519   |     -      |     -     |   7.22   \n",
            "   1    |  2860   |   0.001221   |     -      |     -     |   7.19   \n",
            "   1    |  2880   |   0.010844   |     -      |     -     |   7.22   \n",
            "   1    |  2900   |   0.001181   |     -      |     -     |   7.22   \n",
            "   1    |  2920   |   0.000720   |     -      |     -     |   7.20   \n",
            "   1    |  2940   |   0.000599   |     -      |     -     |   7.23   \n",
            "   1    |  2960   |   0.000538   |     -      |     -     |   7.24   \n",
            "   1    |  2980   |   0.000501   |     -      |     -     |   7.22   \n",
            "   1    |  3000   |   0.010676   |     -      |     -     |   7.21   \n",
            "   1    |  3020   |   0.012471   |     -      |     -     |   7.21   \n",
            "   1    |  3040   |   0.008870   |     -      |     -     |   7.22   \n",
            "   1    |  3060   |   0.012615   |     -      |     -     |   7.23   \n",
            "   1    |  3080   |   0.023643   |     -      |     -     |   7.22   \n",
            "   1    |  3100   |   0.000984   |     -      |     -     |   7.24   \n",
            "   1    |  3120   |   0.000798   |     -      |     -     |   7.23   \n",
            "   1    |  3124   |   0.000627   |     -      |     -     |   1.44   \n",
            "----------------------------------------------------------------------\n",
            "\n",
            "\n",
            " Epoch  |  Batch  |  Train Loss  |  Val Loss  |  Val Acc  |  Elapsed \n",
            "----------------------------------------------------------------------\n",
            "   2    |   20    |   0.022990   |     -      |     -     |   7.58   \n",
            "   2    |   40    |   0.000970   |     -      |     -     |   7.22   \n",
            "   2    |   60    |   0.000841   |     -      |     -     |   7.24   \n",
            "   2    |   80    |   0.011839   |     -      |     -     |   7.20   \n",
            "   2    |   100   |   0.000971   |     -      |     -     |   7.26   \n",
            "   2    |   120   |   0.008990   |     -      |     -     |   7.21   \n",
            "   2    |   140   |   0.006152   |     -      |     -     |   7.21   \n",
            "   2    |   160   |   0.013015   |     -      |     -     |   7.23   \n",
            "   2    |   180   |   0.012644   |     -      |     -     |   7.20   \n",
            "   2    |   200   |   0.012327   |     -      |     -     |   7.25   \n",
            "   2    |   220   |   0.001288   |     -      |     -     |   7.20   \n",
            "   2    |   240   |   0.000724   |     -      |     -     |   7.21   \n",
            "   2    |   260   |   0.012261   |     -      |     -     |   7.24   \n",
            "   2    |   280   |   0.000745   |     -      |     -     |   7.22   \n",
            "   2    |   300   |   0.040280   |     -      |     -     |   7.22   \n",
            "   2    |   320   |   0.010619   |     -      |     -     |   7.24   \n",
            "   2    |   340   |   0.005893   |     -      |     -     |   7.26   \n",
            "   2    |   360   |   0.001003   |     -      |     -     |   7.22   \n",
            "   2    |   380   |   0.000888   |     -      |     -     |   7.24   \n",
            "   2    |   400   |   0.014687   |     -      |     -     |   7.21   \n",
            "   2    |   420   |   0.038519   |     -      |     -     |   7.22   \n",
            "   2    |   440   |   0.007500   |     -      |     -     |   7.24   \n",
            "   2    |   460   |   0.001080   |     -      |     -     |   7.22   \n",
            "   2    |   480   |   0.014418   |     -      |     -     |   7.23   \n",
            "   2    |   500   |   0.004749   |     -      |     -     |   7.21   \n",
            "   2    |   520   |   0.012639   |     -      |     -     |   7.21   \n",
            "   2    |   540   |   0.004894   |     -      |     -     |   7.22   \n",
            "   2    |   560   |   0.000615   |     -      |     -     |   7.23   \n",
            "   2    |   580   |   0.001139   |     -      |     -     |   7.22   \n",
            "   2    |   600   |   0.000647   |     -      |     -     |   7.20   \n",
            "   2    |   620   |   0.000507   |     -      |     -     |   7.23   \n",
            "   2    |   640   |   0.012374   |     -      |     -     |   7.19   \n",
            "   2    |   660   |   0.022777   |     -      |     -     |   7.21   \n",
            "   2    |   680   |   0.011626   |     -      |     -     |   7.20   \n",
            "   2    |   700   |   0.000947   |     -      |     -     |   7.22   \n",
            "   2    |   720   |   0.012308   |     -      |     -     |   7.23   \n",
            "   2    |   740   |   0.028063   |     -      |     -     |   7.22   \n",
            "   2    |   760   |   0.001551   |     -      |     -     |   7.23   \n",
            "   2    |   780   |   0.004883   |     -      |     -     |   7.21   \n",
            "   2    |   800   |   0.012497   |     -      |     -     |   7.23   \n",
            "   2    |   820   |   0.000969   |     -      |     -     |   7.20   \n",
            "   2    |   840   |   0.031218   |     -      |     -     |   7.21   \n",
            "   2    |   860   |   0.009289   |     -      |     -     |   7.22   \n",
            "   2    |   880   |   0.001375   |     -      |     -     |   7.23   \n",
            "   2    |   900   |   0.000705   |     -      |     -     |   7.23   \n",
            "   2    |   920   |   0.019816   |     -      |     -     |   7.25   \n",
            "   2    |   940   |   0.001276   |     -      |     -     |   7.21   \n",
            "   2    |   960   |   0.012578   |     -      |     -     |   7.23   \n",
            "   2    |   980   |   0.012329   |     -      |     -     |   7.22   \n",
            "   2    |  1000   |   0.011071   |     -      |     -     |   7.27   \n",
            "   2    |  1020   |   0.001159   |     -      |     -     |   7.25   \n",
            "   2    |  1040   |   0.018288   |     -      |     -     |   7.24   \n",
            "   2    |  1060   |   0.001365   |     -      |     -     |   7.20   \n",
            "   2    |  1080   |   0.004664   |     -      |     -     |   7.20   \n",
            "   2    |  1100   |   0.010618   |     -      |     -     |   7.21   \n",
            "   2    |  1120   |   0.000761   |     -      |     -     |   7.23   \n",
            "   2    |  1140   |   0.004660   |     -      |     -     |   7.21   \n",
            "   2    |  1160   |   0.000985   |     -      |     -     |   7.22   \n",
            "   2    |  1180   |   0.012541   |     -      |     -     |   7.21   \n",
            "   2    |  1200   |   0.000549   |     -      |     -     |   7.19   \n",
            "   2    |  1220   |   0.000807   |     -      |     -     |   7.22   \n",
            "   2    |  1240   |   0.000987   |     -      |     -     |   7.23   \n",
            "   2    |  1260   |   0.000534   |     -      |     -     |   7.20   \n",
            "   2    |  1280   |   0.000702   |     -      |     -     |   7.23   \n",
            "   2    |  1300   |   0.000464   |     -      |     -     |   7.21   \n",
            "   2    |  1320   |   0.000436   |     -      |     -     |   7.22   \n",
            "   2    |  1340   |   0.013356   |     -      |     -     |   7.23   \n",
            "   2    |  1360   |   0.000455   |     -      |     -     |   7.25   \n",
            "   2    |  1380   |   0.000570   |     -      |     -     |   7.21   \n",
            "   2    |  1400   |   0.012863   |     -      |     -     |   7.23   \n",
            "   2    |  1420   |   0.000503   |     -      |     -     |   7.21   \n",
            "   2    |  1440   |   0.018297   |     -      |     -     |   7.24   \n",
            "   2    |  1460   |   0.000422   |     -      |     -     |   7.21   \n",
            "   2    |  1480   |   0.000506   |     -      |     -     |   7.22   \n",
            "   2    |  1500   |   0.000474   |     -      |     -     |   7.25   \n",
            "   2    |  1520   |   0.012732   |     -      |     -     |   7.19   \n",
            "   2    |  1540   |   0.000427   |     -      |     -     |   7.20   \n",
            "   2    |  1560   |   0.000697   |     -      |     -     |   7.22   \n",
            "   2    |  1580   |   0.024847   |     -      |     -     |   7.33   \n",
            "   2    |  1600   |   0.012152   |     -      |     -     |   7.20   \n",
            "   2    |  1620   |   0.000563   |     -      |     -     |   7.24   \n",
            "   2    |  1640   |   0.005588   |     -      |     -     |   7.17   \n",
            "   2    |  1660   |   0.009096   |     -      |     -     |   7.18   \n",
            "   2    |  1680   |   0.012495   |     -      |     -     |   7.20   \n",
            "   2    |  1700   |   0.000874   |     -      |     -     |   7.19   \n",
            "   2    |  1720   |   0.000943   |     -      |     -     |   7.19   \n",
            "   2    |  1740   |   0.009066   |     -      |     -     |   7.19   \n",
            "   2    |  1760   |   0.012077   |     -      |     -     |   7.19   \n",
            "   2    |  1780   |   0.023658   |     -      |     -     |   7.20   \n",
            "   2    |  1800   |   0.001326   |     -      |     -     |   7.19   \n",
            "   2    |  1820   |   0.012572   |     -      |     -     |   7.19   \n",
            "   2    |  1840   |   0.007454   |     -      |     -     |   7.17   \n",
            "   2    |  1860   |   0.001442   |     -      |     -     |   7.21   \n",
            "   2    |  1880   |   0.000846   |     -      |     -     |   7.21   \n",
            "   2    |  1900   |   0.009038   |     -      |     -     |   7.20   \n",
            "   2    |  1920   |   0.020242   |     -      |     -     |   7.20   \n",
            "   2    |  1940   |   0.000687   |     -      |     -     |   7.19   \n",
            "   2    |  1960   |   0.008578   |     -      |     -     |   7.21   \n",
            "   2    |  1980   |   0.011206   |     -      |     -     |   7.18   \n",
            "   2    |  2000   |   0.004491   |     -      |     -     |   7.19   \n",
            "   2    |  2020   |   0.000737   |     -      |     -     |   7.19   \n",
            "   2    |  2040   |   0.012683   |     -      |     -     |   7.21   \n",
            "   2    |  2060   |   0.012603   |     -      |     -     |   7.22   \n",
            "   2    |  2080   |   0.000814   |     -      |     -     |   7.20   \n",
            "   2    |  2100   |   0.037734   |     -      |     -     |   7.21   \n",
            "   2    |  2120   |   0.008328   |     -      |     -     |   7.19   \n",
            "   2    |  2140   |   0.037358   |     -      |     -     |   7.21   \n",
            "   2    |  2160   |   0.001312   |     -      |     -     |   7.22   \n",
            "   2    |  2180   |   0.013159   |     -      |     -     |   7.24   \n",
            "   2    |  2200   |   0.000985   |     -      |     -     |   7.24   \n",
            "   2    |  2220   |   0.001007   |     -      |     -     |   7.21   \n",
            "   2    |  2240   |   0.004926   |     -      |     -     |   7.23   \n",
            "   2    |  2260   |   0.000876   |     -      |     -     |   7.18   \n",
            "   2    |  2280   |   0.000837   |     -      |     -     |   7.22   \n",
            "   2    |  2300   |   0.000866   |     -      |     -     |   7.23   \n",
            "   2    |  2320   |   0.000728   |     -      |     -     |   7.21   \n",
            "   2    |  2340   |   0.024232   |     -      |     -     |   7.20   \n",
            "   2    |  2360   |   0.025788   |     -      |     -     |   7.19   \n",
            "   2    |  2380   |   0.000921   |     -      |     -     |   7.21   \n",
            "   2    |  2400   |   0.006085   |     -      |     -     |   7.23   \n",
            "   2    |  2420   |   0.010994   |     -      |     -     |   7.17   \n",
            "   2    |  2440   |   0.003466   |     -      |     -     |   7.21   \n",
            "   2    |  2460   |   0.024403   |     -      |     -     |   7.24   \n",
            "   2    |  2480   |   0.005195   |     -      |     -     |   7.21   \n",
            "   2    |  2500   |   0.001258   |     -      |     -     |   7.17   \n",
            "   2    |  2520   |   0.028269   |     -      |     -     |   7.20   \n",
            "   2    |  2540   |   0.000803   |     -      |     -     |   7.19   \n",
            "   2    |  2560   |   0.000778   |     -      |     -     |   7.18   \n",
            "   2    |  2580   |   0.010811   |     -      |     -     |   7.17   \n",
            "   2    |  2600   |   0.001177   |     -      |     -     |   7.18   \n",
            "   2    |  2620   |   0.014666   |     -      |     -     |   7.20   \n",
            "   2    |  2640   |   0.003899   |     -      |     -     |   7.17   \n",
            "   2    |  2660   |   0.001485   |     -      |     -     |   7.20   \n",
            "   2    |  2680   |   0.012420   |     -      |     -     |   7.16   \n",
            "   2    |  2700   |   0.020734   |     -      |     -     |   7.19   \n",
            "   2    |  2720   |   0.011260   |     -      |     -     |   7.17   \n",
            "   2    |  2740   |   0.000923   |     -      |     -     |   7.18   \n",
            "   2    |  2760   |   0.024144   |     -      |     -     |   7.17   \n",
            "   2    |  2780   |   0.001062   |     -      |     -     |   7.17   \n",
            "   2    |  2800   |   0.001010   |     -      |     -     |   7.19   \n",
            "   2    |  2820   |   0.000850   |     -      |     -     |   7.18   \n",
            "   2    |  2840   |   0.008897   |     -      |     -     |   7.17   \n",
            "   2    |  2860   |   0.012521   |     -      |     -     |   7.18   \n",
            "   2    |  2880   |   0.006813   |     -      |     -     |   7.18   \n",
            "   2    |  2900   |   0.000800   |     -      |     -     |   7.19   \n",
            "   2    |  2920   |   0.004011   |     -      |     -     |   7.19   \n",
            "   2    |  2940   |   0.000876   |     -      |     -     |   7.18   \n",
            "   2    |  2960   |   0.003622   |     -      |     -     |   7.20   \n",
            "   2    |  2980   |   0.001405   |     -      |     -     |   7.22   \n",
            "   2    |  3000   |   0.013007   |     -      |     -     |   7.20   \n",
            "   2    |  3020   |   0.012269   |     -      |     -     |   7.21   \n",
            "   2    |  3040   |   0.027121   |     -      |     -     |   7.20   \n",
            "   2    |  3060   |   0.000874   |     -      |     -     |   7.22   \n",
            "   2    |  3080   |   0.001036   |     -      |     -     |   7.23   \n",
            "   2    |  3100   |   0.012524   |     -      |     -     |   7.22   \n",
            "   2    |  3120   |   0.012780   |     -      |     -     |   7.22   \n",
            "   2    |  3124   |   0.000926   |     -      |     -     |   1.44   \n",
            "----------------------------------------------------------------------\n",
            "\n",
            "\n",
            "Training complete!\n"
          ]
        }
      ],
      "source": [
        "# Concatenate the train set and the validation set\n",
        "full_train_data = torch.utils.data.ConcatDataset([train_data, val_data])\n",
        "full_train_sampler = RandomSampler(full_train_data)\n",
        "full_train_dataloader = DataLoader(full_train_data, sampler=full_train_sampler, batch_size=32)\n",
        "\n",
        "# Train the Bert Classifier on the entire training data\n",
        "set_seed(42)\n",
        "bert_classifier, optimizer, scheduler = initialize_model(epochs=2)\n",
        "train(bert_classifier, full_train_dataloader, epochs=2)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 63,
      "metadata": {
        "id": "LZUnDt4x-E-2",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 206
        },
        "outputId": "bab11935-0c00-400a-cdb2-89a12f8c2f45"
      },
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "            id                                             review\n",
              "660   46421986  I booked and they accepted the reservation and...\n",
              "1353  12856992  Uğur's home was very clean and safely. People'...\n",
              "861   31050262                            lugar sólo para mujeres\n",
              "111   22026830  The apartment looks even better than it does i...\n",
              "332   51493500  I got a confirmation twice and this host cance..."
            ],
            "text/html": [
              "\n",
              "  <div id=\"df-89bbc561-0dfc-44e0-b451-9b3b02aa1fcd\" class=\"colab-df-container\">\n",
              "    <div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>id</th>\n",
              "      <th>review</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>660</th>\n",
              "      <td>46421986</td>\n",
              "      <td>I booked and they accepted the reservation and...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1353</th>\n",
              "      <td>12856992</td>\n",
              "      <td>Uğur's home was very clean and safely. People'...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>861</th>\n",
              "      <td>31050262</td>\n",
              "      <td>lugar sólo para mujeres</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>111</th>\n",
              "      <td>22026830</td>\n",
              "      <td>The apartment looks even better than it does i...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>332</th>\n",
              "      <td>51493500</td>\n",
              "      <td>I got a confirmation twice and this host cance...</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>\n",
              "    <div class=\"colab-df-buttons\">\n",
              "\n",
              "  <div class=\"colab-df-container\">\n",
              "    <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-89bbc561-0dfc-44e0-b451-9b3b02aa1fcd')\"\n",
              "            title=\"Convert this dataframe to an interactive table.\"\n",
              "            style=\"display:none;\">\n",
              "\n",
              "  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\" viewBox=\"0 -960 960 960\">\n",
              "    <path d=\"M120-120v-720h720v720H120Zm60-500h600v-160H180v160Zm220 220h160v-160H400v160Zm0 220h160v-160H400v160ZM180-400h160v-160H180v160Zm440 0h160v-160H620v160ZM180-180h160v-160H180v160Zm440 0h160v-160H620v160Z\"/>\n",
              "  </svg>\n",
              "    </button>\n",
              "\n",
              "  <style>\n",
              "    .colab-df-container {\n",
              "      display:flex;\n",
              "      gap: 12px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert {\n",
              "      background-color: #E8F0FE;\n",
              "      border: none;\n",
              "      border-radius: 50%;\n",
              "      cursor: pointer;\n",
              "      display: none;\n",
              "      fill: #1967D2;\n",
              "      height: 32px;\n",
              "      padding: 0 0 0 0;\n",
              "      width: 32px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert:hover {\n",
              "      background-color: #E2EBFA;\n",
              "      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
              "      fill: #174EA6;\n",
              "    }\n",
              "\n",
              "    .colab-df-buttons div {\n",
              "      margin-bottom: 4px;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert {\n",
              "      background-color: #3B4455;\n",
              "      fill: #D2E3FC;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert:hover {\n",
              "      background-color: #434B5C;\n",
              "      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
              "      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
              "      fill: #FFFFFF;\n",
              "    }\n",
              "  </style>\n",
              "\n",
              "    <script>\n",
              "      const buttonEl =\n",
              "        document.querySelector('#df-89bbc561-0dfc-44e0-b451-9b3b02aa1fcd button.colab-df-convert');\n",
              "      buttonEl.style.display =\n",
              "        google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "\n",
              "      async function convertToInteractive(key) {\n",
              "        const element = document.querySelector('#df-89bbc561-0dfc-44e0-b451-9b3b02aa1fcd');\n",
              "        const dataTable =\n",
              "          await google.colab.kernel.invokeFunction('convertToInteractive',\n",
              "                                                    [key], {});\n",
              "        if (!dataTable) return;\n",
              "\n",
              "        const docLinkHtml = 'Like what you see? Visit the ' +\n",
              "          '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n",
              "          + ' to learn more about interactive tables.';\n",
              "        element.innerHTML = '';\n",
              "        dataTable['output_type'] = 'display_data';\n",
              "        await google.colab.output.renderOutput(dataTable, element);\n",
              "        const docLink = document.createElement('div');\n",
              "        docLink.innerHTML = docLinkHtml;\n",
              "        element.appendChild(docLink);\n",
              "      }\n",
              "    </script>\n",
              "  </div>\n",
              "\n",
              "\n",
              "<div id=\"df-31d83e67-f232-4a35-8c1e-c9c5662bb2eb\">\n",
              "  <button class=\"colab-df-quickchart\" onclick=\"quickchart('df-31d83e67-f232-4a35-8c1e-c9c5662bb2eb')\"\n",
              "            title=\"Suggest charts\"\n",
              "            style=\"display:none;\">\n",
              "\n",
              "<svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n",
              "     width=\"24px\">\n",
              "    <g>\n",
              "        <path d=\"M19 3H5c-1.1 0-2 .9-2 2v14c0 1.1.9 2 2 2h14c1.1 0 2-.9 2-2V5c0-1.1-.9-2-2-2zM9 17H7v-7h2v7zm4 0h-2V7h2v10zm4 0h-2v-4h2v4z\"/>\n",
              "    </g>\n",
              "</svg>\n",
              "  </button>\n",
              "\n",
              "<style>\n",
              "  .colab-df-quickchart {\n",
              "      --bg-color: #E8F0FE;\n",
              "      --fill-color: #1967D2;\n",
              "      --hover-bg-color: #E2EBFA;\n",
              "      --hover-fill-color: #174EA6;\n",
              "      --disabled-fill-color: #AAA;\n",
              "      --disabled-bg-color: #DDD;\n",
              "  }\n",
              "\n",
              "  [theme=dark] .colab-df-quickchart {\n",
              "      --bg-color: #3B4455;\n",
              "      --fill-color: #D2E3FC;\n",
              "      --hover-bg-color: #434B5C;\n",
              "      --hover-fill-color: #FFFFFF;\n",
              "      --disabled-bg-color: #3B4455;\n",
              "      --disabled-fill-color: #666;\n",
              "  }\n",
              "\n",
              "  .colab-df-quickchart {\n",
              "    background-color: var(--bg-color);\n",
              "    border: none;\n",
              "    border-radius: 50%;\n",
              "    cursor: pointer;\n",
              "    display: none;\n",
              "    fill: var(--fill-color);\n",
              "    height: 32px;\n",
              "    padding: 0;\n",
              "    width: 32px;\n",
              "  }\n",
              "\n",
              "  .colab-df-quickchart:hover {\n",
              "    background-color: var(--hover-bg-color);\n",
              "    box-shadow: 0 1px 2px rgba(60, 64, 67, 0.3), 0 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
              "    fill: var(--button-hover-fill-color);\n",
              "  }\n",
              "\n",
              "  .colab-df-quickchart-complete:disabled,\n",
              "  .colab-df-quickchart-complete:disabled:hover {\n",
              "    background-color: var(--disabled-bg-color);\n",
              "    fill: var(--disabled-fill-color);\n",
              "    box-shadow: none;\n",
              "  }\n",
              "\n",
              "  .colab-df-spinner {\n",
              "    border: 2px solid var(--fill-color);\n",
              "    border-color: transparent;\n",
              "    border-bottom-color: var(--fill-color);\n",
              "    animation:\n",
              "      spin 1s steps(1) infinite;\n",
              "  }\n",
              "\n",
              "  @keyframes spin {\n",
              "    0% {\n",
              "      border-color: transparent;\n",
              "      border-bottom-color: var(--fill-color);\n",
              "      border-left-color: var(--fill-color);\n",
              "    }\n",
              "    20% {\n",
              "      border-color: transparent;\n",
              "      border-left-color: var(--fill-color);\n",
              "      border-top-color: var(--fill-color);\n",
              "    }\n",
              "    30% {\n",
              "      border-color: transparent;\n",
              "      border-left-color: var(--fill-color);\n",
              "      border-top-color: var(--fill-color);\n",
              "      border-right-color: var(--fill-color);\n",
              "    }\n",
              "    40% {\n",
              "      border-color: transparent;\n",
              "      border-right-color: var(--fill-color);\n",
              "      border-top-color: var(--fill-color);\n",
              "    }\n",
              "    60% {\n",
              "      border-color: transparent;\n",
              "      border-right-color: var(--fill-color);\n",
              "    }\n",
              "    80% {\n",
              "      border-color: transparent;\n",
              "      border-right-color: var(--fill-color);\n",
              "      border-bottom-color: var(--fill-color);\n",
              "    }\n",
              "    90% {\n",
              "      border-color: transparent;\n",
              "      border-bottom-color: var(--fill-color);\n",
              "    }\n",
              "  }\n",
              "</style>\n",
              "\n",
              "  <script>\n",
              "    async function quickchart(key) {\n",
              "      const quickchartButtonEl =\n",
              "        document.querySelector('#' + key + ' button');\n",
              "      quickchartButtonEl.disabled = true;  // To prevent multiple clicks.\n",
              "      quickchartButtonEl.classList.add('colab-df-spinner');\n",
              "      try {\n",
              "        const charts = await google.colab.kernel.invokeFunction(\n",
              "            'suggestCharts', [key], {});\n",
              "      } catch (error) {\n",
              "        console.error('Error during call to suggestCharts:', error);\n",
              "      }\n",
              "      quickchartButtonEl.classList.remove('colab-df-spinner');\n",
              "      quickchartButtonEl.classList.add('colab-df-quickchart-complete');\n",
              "    }\n",
              "    (() => {\n",
              "      let quickchartButtonEl =\n",
              "        document.querySelector('#df-31d83e67-f232-4a35-8c1e-c9c5662bb2eb button');\n",
              "      quickchartButtonEl.style.display =\n",
              "        google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "    })();\n",
              "  </script>\n",
              "</div>\n",
              "    </div>\n",
              "  </div>\n"
            ]
          },
          "metadata": {},
          "execution_count": 63
        }
      ],
      "source": [
        "test_data.sample(5)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 64,
      "metadata": {
        "id": "L7rKCAtQ-E-2",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "e6f0ce70-bd65-4e24-a139-9ffd4df15f64"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Tokenizing data...\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/transformers/tokenization_utils_base.py:2614: FutureWarning: The `pad_to_max_length` argument is deprecated and will be removed in a future version, use `padding=True` or `padding='longest'` to pad to the longest sequence in the batch, or use `padding='max_length'` to pad to a max length. In this case, you can give a specific length with `max_length` (e.g. `max_length=45`) or leave max_length to None to pad to the maximal input size of the model (e.g. 512 for Bert).\n",
            "  warnings.warn(\n"
          ]
        }
      ],
      "source": [
        "# Run `preprocessing_for_bert` on the test set\n",
        "print('Tokenizing data...')\n",
        "test_inputs, test_masks = preprocessing_for_bert(test_data.review)\n",
        "\n",
        "# Create the DataLoader for our test set\n",
        "test_dataset = TensorDataset(test_inputs, test_masks)\n",
        "test_sampler = SequentialSampler(test_dataset)\n",
        "test_dataloader = DataLoader(test_dataset, sampler=test_sampler, batch_size=32)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 86,
      "metadata": {
        "id": "0OEGtAik-E-2",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "09817424-744e-4d95-8527-99fdc1656001"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Number of reviews predicted non-negative:  295\n"
          ]
        }
      ],
      "source": [
        "# Compute predicted probabilities on the test set\n",
        "probs = bert_predict(bert_classifier, test_dataloader)\n",
        "\n",
        "# Get predictions from the probabilities\n",
        "threshold = 0.999604  # manually adjusted to get predicted reviews = 300\n",
        "preds = np.where(probs[:, 1] > threshold, 1, 0)\n",
        "\n",
        "# Number of reviews predicted non-negative\n",
        "print(\"Number of reviews predicted non-negative: \", preds.sum())"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 93,
      "metadata": {
        "id": "yIowZV74-E-3",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "c6bf5992-731a-4e39-df47-b24646fc548a"
      },
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "['This is a great NY find. It is really near to the MTA and had access to the whole city. They building is run down NY but the room is clean and fine. It slept our family of 7 just fine. We would stay here again.',\n",
              " \"Good location and quiet. Myriam she's kind and always helpful person; when ever we needed her, she was responding even late night\",\n",
              " 'AMAZING stay! This is our second time booking with Jose the host is super friendly and responsive,the place is always super clean and super comfortable. You get your own private bathroom as well which is very hard to find on bnb to begin with when booking private rooms along with a fully functioning shared kitchen and the neighborhood is very safe. Would recommend 1000%!',\n",
              " 'We were in town for work and we were able to enjoy our stay. Working nights and sleeping in the day made being able to rest in the daytime very important. We enjoyed a quiet peaceful stay, thank you.',\n",
              " 'Very clean/quiet place. The sheets were clean and did not have any weird smell. The room temperature was perfect and comfortable. Michelle is a very responsive host. Will stay here again!',\n",
              " 'Cute, clean apartment in a decent neighborhood with a great nightlife. Experienced a lot of difficulty with the check-in process due to lack of transparency & communication from the host',\n",
              " 'We stayed at Chris’s place for a week and it was great. Perfect location, lots of great food nearby, multiple subway stations are very close by. Safe area to walk at night and the apartment was warm with a comfy bed. Good value for the location we would recommend staying here. Try the ramen shop up the road which is very good. The umbrellas were a nice touch as we had a couple days of rain/snow during our visit.',\n",
              " \"We had a great stay at Christiana's apartment. We were pretty much only there at night to sleep, but it was a comfortable apartment close to both public transportation and a fun Williamsburg bar scene. \",\n",
              " 'Great quiet location close to the train. The space is comfortable, clean and perfect to come back to after a day of sightseeing. We would come back again!',\n",
              " \"Bianca did a great job of setting expectations for the room. With that said, there was more space in the unit than we thought we would have which was great. It's a great space for those who have a busy, short weekend planned in NYC and simply need a place to put there stuff and crash.\",\n",
              " 'best place in bushwick!<br/>lovely neighbourhood, beautiful apartment<br/>cant recommend enough',\n",
              " \"The apartment was very cosy and the hostess was responding very fast when we had questions. The house is very close to the metro station and well equipped. We stronlgy recommend it in case you want to visit New York and don't want to stay in Manhattan.\",\n",
              " 'Great location! Close to some excellent restaurants and right by Times Square. Eran was very responsive as well and quick to answer any questions. Would stay here again!',\n",
              " 'We enjoyed our stay at Elise’s place!  My kids loved all the books and the trampoline!  The beds were comfy.  The kitchen and bathrooms were updated and modern.  Communication was fast and easy.  Thanks!',\n",
              " 'I had a great experience staying at Bassams property. It was very clean and comfortable. Convenient location, close to many restaurants and the R and N trains. I would highly recommend staying here.',\n",
              " 'Luca’s place is cute and comfortable, and really well-located. The couch and bed were comfortable, and the place served us well as we explored Prospect Heights. We had a snafu when checking in, but the team managing the place was very communicative and fast to get to a resolution for us!',\n",
              " 'Definitely enjoyed my stay at Nina’s, Very modern apartment with nice appliances. Shower was amazing. It definitely can get noisy (so keep that in mind) but I was very familiar with the area prior to booking so It didn’t bother me. Would most definitely book again for future stays in NYC. I highly recommend this spot. ☺️',\n",
              " 'Great location but can be very lively because of the traffic and neighborhood noises. The place itself had all the basics and is easily accessible to trains and good restaurants. The bed was a little bit uncomfortable for my preference but overall a decent stay.',\n",
              " 'Laura is so sweet, friendly  and wonderful. great location, nice neighborhood. room is cozy and comfortable. a little bit small and not much sunshine, but fine with me. would like to stay again! Thank you Laura!',\n",
              " 'I totally loved my stayed Mohammed was so kind he even picked me up from the airport! Thank you very much ! the room was very clean and  spacious']"
            ]
          },
          "metadata": {},
          "execution_count": 93
        }
      ],
      "source": [
        "output = test_data[preds==1]\n",
        "list(output.sample(20).review)"
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "torch.save(bert_classifier, '/content/drive/MyDrive/Colab Notebooks/finetuneBERT.ipynb')"
      ],
      "metadata": {
        "id": "8KIKpgNVqYOs"
      },
      "execution_count": 97,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "saved_model = torch.load('/content/drive/MyDrive/Colab Notebooks/finetuneBERT.ipynb')"
      ],
      "metadata": {
        "id": "ZctxFPVMrIbS"
      },
      "execution_count": 100,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "saved_model"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "k0MEbS4VrUqz",
        "outputId": "2e92681a-e7be-4d29-a132-999099a588ff"
      },
      "execution_count": 101,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "BertClassifier(\n",
              "  (bert): BertModel(\n",
              "    (embeddings): BertEmbeddings(\n",
              "      (word_embeddings): Embedding(30522, 768, padding_idx=0)\n",
              "      (position_embeddings): Embedding(512, 768)\n",
              "      (token_type_embeddings): Embedding(2, 768)\n",
              "      (LayerNorm): LayerNorm((768,), eps=1e-12, elementwise_affine=True)\n",
              "      (dropout): Dropout(p=0.1, inplace=False)\n",
              "    )\n",
              "    (encoder): BertEncoder(\n",
              "      (layer): ModuleList(\n",
              "        (0-11): 12 x BertLayer(\n",
              "          (attention): BertAttention(\n",
              "            (self): BertSelfAttention(\n",
              "              (query): Linear(in_features=768, out_features=768, bias=True)\n",
              "              (key): Linear(in_features=768, out_features=768, bias=True)\n",
              "              (value): Linear(in_features=768, out_features=768, bias=True)\n",
              "              (dropout): Dropout(p=0.1, inplace=False)\n",
              "            )\n",
              "            (output): BertSelfOutput(\n",
              "              (dense): Linear(in_features=768, out_features=768, bias=True)\n",
              "              (LayerNorm): LayerNorm((768,), eps=1e-12, elementwise_affine=True)\n",
              "              (dropout): Dropout(p=0.1, inplace=False)\n",
              "            )\n",
              "          )\n",
              "          (intermediate): BertIntermediate(\n",
              "            (dense): Linear(in_features=768, out_features=3072, bias=True)\n",
              "            (intermediate_act_fn): GELUActivation()\n",
              "          )\n",
              "          (output): BertOutput(\n",
              "            (dense): Linear(in_features=3072, out_features=768, bias=True)\n",
              "            (LayerNorm): LayerNorm((768,), eps=1e-12, elementwise_affine=True)\n",
              "            (dropout): Dropout(p=0.1, inplace=False)\n",
              "          )\n",
              "        )\n",
              "      )\n",
              "    )\n",
              "    (pooler): BertPooler(\n",
              "      (dense): Linear(in_features=768, out_features=768, bias=True)\n",
              "      (activation): Tanh()\n",
              "    )\n",
              "  )\n",
              "  (classifier): Sequential(\n",
              "    (0): Linear(in_features=768, out_features=50, bias=True)\n",
              "    (1): ReLU()\n",
              "    (2): Linear(in_features=50, out_features=2, bias=True)\n",
              "  )\n",
              ")"
            ]
          },
          "metadata": {},
          "execution_count": 101
        }
      ]
    }
  ],
  "metadata": {
    "kernelspec": {
      "display_name": "Python 3",
      "name": "python3"
    },
    "language_info": {
      "codemirror_mode": {
        "name": "ipython",
        "version": 3
      },
      "file_extension": ".py",
      "mimetype": "text/x-python",
      "name": "python",
      "nbconvert_exporter": "python",
      "pygments_lexer": "ipython3",
      "version": "3.10.12"
    },
    "colab": {
      "provenance": [],
      "gpuType": "T4",
      "cell_execution_strategy": "setup",
      "include_colab_link": true
    },
    "accelerator": "GPU"
  },
  "nbformat": 4,
  "nbformat_minor": 0
}