GPT3_generate_paraphrases.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "provenance": []
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    },
    "language_info": {
      "name": "python"
    }
  },
  "cells": [
    {
      "cell_type": "markdown",
      "source": [
        "#Installing Dependencies"
      ],
      "metadata": {
        "id": "n2SmYI5usrMa"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "# !pip install openai datasets\n",
        "!pip install openai"
      ],
      "metadata": {
        "id": "SWsg7Mkwwqsh",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "574024c9-553e-4ab0-8d66-51ae94d8e24b"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Looking in indexes: https://pypi.org/simple, https://us-python.pkg.dev/colab-wheels/public/simple/\n",
            "Collecting openai\n",
            "  Downloading openai-0.26.0.tar.gz (54 kB)\n",
            "\u001b[2K     \u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m54.7/54.7 KB\u001b[0m \u001b[31m4.6 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0m\n",
            "\u001b[?25h  Installing build dependencies ... \u001b[?25l\u001b[?25hdone\n",
            "  Getting requirements to build wheel ... \u001b[?25l\u001b[?25hdone\n",
            "  Preparing metadata (pyproject.toml) ... \u001b[?25l\u001b[?25hdone\n",
            "Requirement already satisfied: aiohttp in /usr/local/lib/python3.8/dist-packages (from openai) (3.8.3)\n",
            "Requirement already satisfied: requests>=2.20 in /usr/local/lib/python3.8/dist-packages (from openai) (2.25.1)\n",
            "Requirement already satisfied: tqdm in /usr/local/lib/python3.8/dist-packages (from openai) (4.64.1)\n",
            "Requirement already satisfied: idna<3,>=2.5 in /usr/local/lib/python3.8/dist-packages (from requests>=2.20->openai) (2.10)\n",
            "Requirement already satisfied: urllib3<1.27,>=1.21.1 in /usr/local/lib/python3.8/dist-packages (from requests>=2.20->openai) (1.24.3)\n",
            "Requirement already satisfied: chardet<5,>=3.0.2 in /usr/local/lib/python3.8/dist-packages (from requests>=2.20->openai) (4.0.0)\n",
            "Requirement already satisfied: certifi>=2017.4.17 in /usr/local/lib/python3.8/dist-packages (from requests>=2.20->openai) (2022.12.7)\n",
            "Requirement already satisfied: charset-normalizer<3.0,>=2.0 in /usr/local/lib/python3.8/dist-packages (from aiohttp->openai) (2.1.1)\n",
            "Requirement already satisfied: multidict<7.0,>=4.5 in /usr/local/lib/python3.8/dist-packages (from aiohttp->openai) (6.0.3)\n",
            "Requirement already satisfied: yarl<2.0,>=1.0 in /usr/local/lib/python3.8/dist-packages (from aiohttp->openai) (1.8.2)\n",
            "Requirement already satisfied: aiosignal>=1.1.2 in /usr/local/lib/python3.8/dist-packages (from aiohttp->openai) (1.3.1)\n",
            "Requirement already satisfied: async-timeout<5.0,>=4.0.0a3 in /usr/local/lib/python3.8/dist-packages (from aiohttp->openai) (4.0.2)\n",
            "Requirement already satisfied: attrs>=17.3.0 in /usr/local/lib/python3.8/dist-packages (from aiohttp->openai) (22.2.0)\n",
            "Requirement already satisfied: frozenlist>=1.1.1 in /usr/local/lib/python3.8/dist-packages (from aiohttp->openai) (1.3.3)\n",
            "Building wheels for collected packages: openai\n",
            "  Building wheel for openai (pyproject.toml) ... \u001b[?25l\u001b[?25hdone\n",
            "  Created wheel for openai: filename=openai-0.26.0-py3-none-any.whl size=66855 sha256=7043561d7079be8244458bd85f90f2623a5bc3ad339e359719c7e63a87d2f854\n",
            "  Stored in directory: /root/.cache/pip/wheels/7e/4c/c8/31e9d441bd937e2b9076627465f9db43ab6db40f08aae60b66\n",
            "Successfully built openai\n",
            "Installing collected packages: openai\n",
            "Successfully installed openai-0.26.0\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "!pip install -U nltk"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "xlqdLLxLw6is",
        "outputId": "b5f61abd-773c-43e9-ed76-8836bd1f72ba"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Looking in indexes: https://pypi.org/simple, https://us-python.pkg.dev/colab-wheels/public/simple/\n",
            "Requirement already satisfied: nltk in /usr/local/lib/python3.7/dist-packages (3.7)\n",
            "Requirement already satisfied: regex>=2021.8.3 in /usr/local/lib/python3.7/dist-packages (from nltk) (2022.6.2)\n",
            "Requirement already satisfied: joblib in /usr/local/lib/python3.7/dist-packages (from nltk) (1.2.0)\n",
            "Requirement already satisfied: click in /usr/local/lib/python3.7/dist-packages (from nltk) (7.1.2)\n",
            "Requirement already satisfied: tqdm in /usr/local/lib/python3.7/dist-packages (from nltk) (4.64.1)\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "# Importing Packages"
      ],
      "metadata": {
        "id": "aIq2VYaysx0J"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "import ast \n",
        "import os\n",
        "# import openai\n",
        "import pandas as pd\n",
        "\n",
        "# from datasets import load_metric"
      ],
      "metadata": {
        "id": "tFcxBovvg1q5"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "# Loading Data"
      ],
      "metadata": {
        "id": "TzWjYN-vs1Dh"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "difficult_ncs_df_raw = pd.read_csv('export_set_1.csv')\n",
        "diff_dict = {'nc':[]}\n",
        "\n",
        "for index, row in difficult_ncs_df_raw.iterrows():\n",
        "  nc = row['nc']\n",
        "  diff_dict['nc'].append(nc)\n",
        "\n",
        "diff_df = pd.DataFrame.from_dict(diff_dict)"
      ],
      "metadata": {
        "id": "L5vw2Vmu3hoj"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "train_df_raw = pd.read_csv('train_gold.csv')\n",
        "train_dict = {'nc':[], 'paraphrase':[]}\n",
        "\n",
        "for index, row in train_df_raw.iterrows():\n",
        "  nc = row['w1'] + \" \" + row['w2']\n",
        "  train_dict['nc'].append(nc)\n",
        "  train_dict['paraphrase'].append(row['paraphrase'])\n",
        "\n",
        "train_df = pd.DataFrame.from_dict(train_dict)\n",
        "train_df"
      ],
      "metadata": {
        "id": "aoRiM1MMgYhh",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 424
        },
        "outputId": "f1b37624-4104-4c46-bd17-5e2fb62e25cc"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "                      nc                                         paraphrase\n",
              "0     advertising agency                       agency that does advertising\n",
              "1     advertising agency              agency that does advertising business\n",
              "2     advertising agency               agency that does work in advertising\n",
              "3     advertising agency     agency that does work in the advertising field\n",
              "4     advertising agency  agency that does work in the advertising business\n",
              "...                  ...                                                ...\n",
              "5436           work area                            area in which one works\n",
              "5437           work area                       area in which one is working\n",
              "5438           work area                          area in which work occurs\n",
              "5439           work area                         area in which work happens\n",
              "5440           work area                     area in which work takes place\n",
              "\n",
              "[5441 rows x 2 columns]"
            ],
            "text/html": [
              "\n",
              "  <div id=\"df-ccc41bbb-8d4c-458c-9d06-72b708c76395\">\n",
              "    <div 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>nc</th>\n",
              "      <th>paraphrase</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>advertising agency</td>\n",
              "      <td>agency that does advertising</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>advertising agency</td>\n",
              "      <td>agency that does advertising business</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>advertising agency</td>\n",
              "      <td>agency that does work in advertising</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>advertising agency</td>\n",
              "      <td>agency that does work in the advertising field</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>advertising agency</td>\n",
              "      <td>agency that does work in the advertising business</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>...</th>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>5436</th>\n",
              "      <td>work area</td>\n",
              "      <td>area in which one works</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>5437</th>\n",
              "      <td>work area</td>\n",
              "      <td>area in which one is working</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>5438</th>\n",
              "      <td>work area</td>\n",
              "      <td>area in which work occurs</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>5439</th>\n",
              "      <td>work area</td>\n",
              "      <td>area in which work happens</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>5440</th>\n",
              "      <td>work area</td>\n",
              "      <td>area in which work takes place</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "<p>5441 rows × 2 columns</p>\n",
              "</div>\n",
              "      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-ccc41bbb-8d4c-458c-9d06-72b708c76395')\"\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 0 24 24\"\n",
              "       width=\"24px\">\n",
              "    <path d=\"M0 0h24v24H0V0z\" fill=\"none\"/>\n",
              "    <path d=\"M18.56 5.44l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94zm-11 1L8.5 8.5l.94-2.06 2.06-.94-2.06-.94L8.5 2.5l-.94 2.06-2.06.94zm10 10l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94z\"/><path d=\"M17.41 7.96l-1.37-1.37c-.4-.4-.92-.59-1.43-.59-.52 0-1.04.2-1.43.59L10.3 9.45l-7.72 7.72c-.78.78-.78 2.05 0 2.83L4 21.41c.39.39.9.59 1.41.59.51 0 1.02-.2 1.41-.59l7.78-7.78 2.81-2.81c.8-.78.8-2.07 0-2.86zM5.41 20L4 18.59l7.72-7.72 1.47 1.35L5.41 20z\"/>\n",
              "  </svg>\n",
              "      </button>\n",
              "      \n",
              "  <style>\n",
              "    .colab-df-container {\n",
              "      display:flex;\n",
              "      flex-wrap:wrap;\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",
              "    [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-ccc41bbb-8d4c-458c-9d06-72b708c76395 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-ccc41bbb-8d4c-458c-9d06-72b708c76395');\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",
              "  </div>\n",
              "  "
            ]
          },
          "metadata": {},
          "execution_count": 4
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "def load_test_valid_df(file_path):\n",
        "    df = pd.read_csv(file_path, usecols=['nc','paraphrases'])\n",
        "\n",
        "    for i, row in df.iterrows():\n",
        "        string_literal = df.at[i, 'paraphrases']\n",
        "        df.at[i, 'paraphrases'] = ast.literal_eval(string_literal)\n",
        "\n",
        "    return df\n",
        "\n",
        "test_df = load_test_valid_df('test_df.csv')\n",
        "print(test_df)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "xjsLlzmDsXtX",
        "outputId": "72da6ce6-cc5c-421e-d886-3d80d50a4359"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "                       nc                                        paraphrases\n",
            "0            police court  [court for minor offenders arrested by the pol...\n",
            "1           business unit  [unit of business, unit of a business, unit of...\n",
            "2          signature tune  [tune that someone makes their signature, tune...\n",
            "3         spring sunshine  [sunshine of spring, sunshine of the spring, s...\n",
            "4           reflex action  [action caused by reflex, action caused by a r...\n",
            "..                    ...                                                ...\n",
            "105     middle management  [management at middle level, management at the...\n",
            "106          woman worker  [woman who works, woman who works a job, woman...\n",
            "107  virtuoso performance  [performance by a virtuoso, performance done b...\n",
            "108  union representative  [representative of a union, representative of ...\n",
            "109        pension scheme  [scheme defining a pension, scheme defining a ...\n",
            "\n",
            "[110 rows x 2 columns]\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "valid_df = load_test_valid_df('valid_df.csv')\n",
        "print(valid_df)"
      ],
      "metadata": {
        "id": "mqUydBl_q1DU",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "ebf68e68-33f9-461e-b9bf-e381738560e5"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "                            nc  \\\n",
            "0             daytime activity   \n",
            "1               welfare system   \n",
            "2              washing machine   \n",
            "3                  mining area   \n",
            "4              state socialism   \n",
            "5               amateur record   \n",
            "6                   ruler line   \n",
            "7                 sport editor   \n",
            "8                   air filter   \n",
            "9                   love scene   \n",
            "10     interrogation technique   \n",
            "11           publicity purpose   \n",
            "12              property price   \n",
            "13                  drug baron   \n",
            "14              summer meeting   \n",
            "15                  place card   \n",
            "16             transfer market   \n",
            "17              village school   \n",
            "18        fertility counsellor   \n",
            "19             coalition force   \n",
            "20               gene mutation   \n",
            "21            transaction cost   \n",
            "22                cabinet room   \n",
            "23                 home market   \n",
            "24                  water bath   \n",
            "25                 drawing pin   \n",
            "26  telecommunications licence   \n",
            "27           election campaign   \n",
            "\n",
            "                                          paraphrases  \n",
            "0   [activity done in daytime, activity enoyed in ...  \n",
            "1   [system controlling the welfare program, syste...  \n",
            "2   [machine used for washing, machine used for wa...  \n",
            "3   [area for mining, area where mining is done, a...  \n",
            "4   [socialism of a state, socialism of the state,...  \n",
            "5   [record made by an amateur, record set by an a...  \n",
            "6   [line drawn with a ruler, line made with a rul...  \n",
            "7   [editor of sports books, editor of sports broa...  \n",
            "8   [filter for air, filter for the air, filter fo...  \n",
            "9   [scene about love, scene involving love, scene...  \n",
            "10  [technique used in interrogation, technique us...  \n",
            "11  [something whose purpose is to create publicit...  \n",
            "12  [price of property, price of a property, sale ...  \n",
            "13  [baron who sells drugs, baron who deals drugs,...  \n",
            "14  [annual meeting in summer, annual meeting in t...  \n",
            "15  [card at ones place, card specifying one's pla...  \n",
            "16  [market dealing with the transfer of players, ...  \n",
            "17  [school in a village, local school in a villag...  \n",
            "18  [counsellor for fertility advice, counsellor f...  \n",
            "19  [force built by coalition, force made by coali...  \n",
            "20  [mutation of a gene, mutation of the genes, mu...  \n",
            "21  [cost of a transaction, cost of an individual ...  \n",
            "22  [room for the cabinet, room for a cabinet, roo...  \n",
            "23  [market for purchasing homes, market for buyin...  \n",
            "24  [bath filled with water, bath full of water, b...  \n",
            "25  [pin used to fasten a drawing to something, pi...  \n",
            "26  [licence for telecommunications, licence allow...  \n",
            "27  [campaign for votes during an election, campai...  \n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "countparas = 0\n",
        "for i, row in test_df.iterrows():\n",
        "  countparas += len(row['paraphrases'])\n",
        "\n",
        "print(countparas)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "bmGz-bsUbVCO",
        "outputId": "3b47a42a-470d-43a2-b333-cbd44cf9d826"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "4820\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "countparas = 0\n",
        "for i, row in valid_df.iterrows():\n",
        "  countparas += len(row['paraphrases'])\n",
        "\n",
        "print(countparas)"
      ],
      "metadata": {
        "id": "m_KeCC6jbhmX",
        "outputId": "379b227e-ee86-4615-eaa1-79e05958fb34",
        "colab": {
          "base_uri": "https://localhost:8080/"
        }
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "1469\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "import pandas as pd\n",
        "\n",
        "test_df = pd.read_csv(\"new_diff_set.csv\", usecols=['nc'])\n",
        "test_df"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 424
        },
        "id": "IA4ovPLTq5q2",
        "outputId": "45962ce3-e418-457b-b89e-c619a7f87e16"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "                      nc\n",
              "0     accounting sleight\n",
              "1    alchemy observation\n",
              "2           art governor\n",
              "3    assault prosecution\n",
              "4     atomization result\n",
              "..                   ...\n",
              "100          voice shake\n",
              "101      wack philosophy\n",
              "102         warrior gang\n",
              "103          weiner race\n",
              "104     wheel efficiency\n",
              "\n",
              "[105 rows x 1 columns]"
            ],
            "text/html": [
              "\n",
              "  <div id=\"df-12e0ad5d-3d62-42c1-8d2f-40a5fa3e6ca7\">\n",
              "    <div 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>nc</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>accounting sleight</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>alchemy observation</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>art governor</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>assault prosecution</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>atomization result</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>...</th>\n",
              "      <td>...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>100</th>\n",
              "      <td>voice shake</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>101</th>\n",
              "      <td>wack philosophy</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>102</th>\n",
              "      <td>warrior gang</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>103</th>\n",
              "      <td>weiner race</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>104</th>\n",
              "      <td>wheel efficiency</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "<p>105 rows × 1 columns</p>\n",
              "</div>\n",
              "      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-12e0ad5d-3d62-42c1-8d2f-40a5fa3e6ca7')\"\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 0 24 24\"\n",
              "       width=\"24px\">\n",
              "    <path d=\"M0 0h24v24H0V0z\" fill=\"none\"/>\n",
              "    <path d=\"M18.56 5.44l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94zm-11 1L8.5 8.5l.94-2.06 2.06-.94-2.06-.94L8.5 2.5l-.94 2.06-2.06.94zm10 10l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94z\"/><path d=\"M17.41 7.96l-1.37-1.37c-.4-.4-.92-.59-1.43-.59-.52 0-1.04.2-1.43.59L10.3 9.45l-7.72 7.72c-.78.78-.78 2.05 0 2.83L4 21.41c.39.39.9.59 1.41.59.51 0 1.02-.2 1.41-.59l7.78-7.78 2.81-2.81c.8-.78.8-2.07 0-2.86zM5.41 20L4 18.59l7.72-7.72 1.47 1.35L5.41 20z\"/>\n",
              "  </svg>\n",
              "      </button>\n",
              "      \n",
              "  <style>\n",
              "    .colab-df-container {\n",
              "      display:flex;\n",
              "      flex-wrap:wrap;\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",
              "    [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-12e0ad5d-3d62-42c1-8d2f-40a5fa3e6ca7 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-12e0ad5d-3d62-42c1-8d2f-40a5fa3e6ca7');\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",
              "  </div>\n",
              "  "
            ]
          },
          "metadata": {},
          "execution_count": 5
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "def get_random_prompt(train_df, test_df, num_egs=1):\n",
        "  test_row = test_df.sample()\n",
        "  test_nc = test_row['nc'].values[0]\n",
        "  test_paras = test_row['paraphrases'].values[0]\n",
        "\n",
        "  prompt = \"\"\n",
        "  \n",
        "  prompt += train_nc_prompt(train_df, num_egs)\n",
        "  prompt += test_nc_prompt(test_nc)\n",
        "  \n",
        "  return prompt, test_nc, test_paras\n",
        "\n",
        "def train_nc_prompt(train_df, num_egs, prompt_text=\"Q: what is the meaning of {}?\\n\"):\n",
        "  prompt = \"\"\n",
        "  for i in range(num_egs):\n",
        "    train_row = train_df.sample()\n",
        "    train_nc = train_row['nc'].values[0]\n",
        "    train_para = train_row['paraphrase'].values[0]\n",
        "\n",
        "    train_q = prompt_text.format(train_nc)\n",
        "    train_a = \"A: \" + train_para + \"\\n\"\n",
        "\n",
        "    prompt += train_q + train_a\n",
        "  return prompt\n",
        "\n",
        "def test_nc_prompt(test_nc, prompt_text=\"Q: what is the meaning of {}?\\n\"):\n",
        "  test_q = prompt_text.format(test_nc)\n",
        "  test_a = \"A:\"\n",
        "  return test_q + test_a\n",
        "\n",
        "def get_all_test_prompts(train_df, test_df, num_egs=10, prompt_text=\"Q: what is the meaning of {}?\\n\"):\n",
        "  prompts = []\n",
        "  for i, test_row in test_df.iterrows():\n",
        "    test_nc = test_row['nc']\n",
        "    gold_paras = test_row['paraphrases']\n",
        "\n",
        "    prompt = \"\"\n",
        "    prompt += train_nc_prompt(train_df, num_egs, prompt_text)\n",
        "    prompt += test_nc_prompt(test_nc, prompt_text)\n",
        "\n",
        "    prompts.append((test_nc, gold_paras, prompt))\n",
        "  \n",
        "  return prompts\n",
        "\n",
        "def get_all_diff_prompts(train_df, diff_df, num_egs=10, prompt_text=\"Q: what is the meaning of {}?\\n\"):\n",
        "  prompts = []\n",
        "  for i, row in diff_df.iterrows():\n",
        "    diff_nc = row['nc']\n",
        "\n",
        "    prompt = \"\"\n",
        "    prompt += train_nc_prompt(train_df, num_egs, prompt_text)\n",
        "    prompt += test_nc_prompt(diff_nc, prompt_text)\n",
        "\n",
        "    prompts.append((diff_nc, prompt))\n",
        "  \n",
        "  return prompts"
      ],
      "metadata": {
        "id": "QGR2kVHliCsQ"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "prompt = train_nc_prompt(train_df, 10) + test_nc_prompt(\"graduation parking\")\n",
        "print(prompt)"
      ],
      "metadata": {
        "id": "INsWHSGjk0xp",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "5879b610-e16a-4681-d5e3-3046af2531aa"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Q: what is the meaning of history book?\n",
            "A: book with history\n",
            "Q: what is the meaning of household account?\n",
            "A: investment account for an entire household\n",
            "Q: what is the meaning of police spokesman?\n",
            "A: spokesman who speaks on the behalf of the police\n",
            "Q: what is the meaning of bank deposit?\n",
            "A: deposit at a bank\n",
            "Q: what is the meaning of corporation tax?\n",
            "A: tax that corporations are required to pay\n",
            "Q: what is the meaning of divorce rate?\n",
            "A: rate of divorce\n",
            "Q: what is the meaning of telephone conversation?\n",
            "A: conversation held through a telephone\n",
            "Q: what is the meaning of census record?\n",
            "A: record of a census\n",
            "Q: what is the meaning of music television?\n",
            "A: television channel where music is shown\n",
            "Q: what is the meaning of fire officer?\n",
            "A: officer who fights fire\n",
            "Q: what is the meaning of graduation parking?\n",
            "A:\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "openai.organization = \"<ORGANIZATION-KEY>\"\n",
        "openai.api_key = \"<API-KEY>\""
      ],
      "metadata": {
        "id": "zELv8efUi7TU"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "# difficult NC generation\n",
        "import time\n",
        "\n",
        "num_egs = 1\n",
        "all_prompts = get_all_diff_prompts(train_df, test_df, num_egs, \"Q: what is the meaning of {}?\\n\")\n",
        "\n",
        "text_responses = {}\n",
        "curr_index = 0\n",
        "\n",
        "for diff_nc, prompt in all_prompts[curr_index:]:\n",
        "  # generate number of responses equal to the number of gold paraphrases\n",
        "  time.sleep(15)\n",
        "  num_to_gen = 5\n",
        "  generated_paras = []\n",
        "  responses = openai.Completion.create(\n",
        "    model=\"text-davinci-002\",\n",
        "    prompt=prompt,\n",
        "    temperature=1,\n",
        "    max_tokens=256,\n",
        "    top_p=1,\n",
        "    frequency_penalty=0,\n",
        "    presence_penalty=0,\n",
        "    n=num_to_gen\n",
        "  )\n",
        "  for r in responses['choices']:\n",
        "    generated_paras.append(r.text.lstrip())\n",
        "\n",
        "  curr_index += 1\n",
        "  print(\"'{}': '{}'\".format(diff_nc, generated_paras[0]))\n",
        "  \n",
        "  text_responses[diff_nc] = generated_paras"
      ],
      "metadata": {
        "id": "ImBNtW6X5b-o"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "import time\n",
        "\n",
        "num_egs = 1\n",
        "all_prompts = get_all_test_prompts(train_df, test_df, num_egs, \"Q: what is the meaning of {}?\\n\")\n",
        "\n",
        "text_responses = {}\n",
        "curr_index = 0\n",
        "\n",
        "# for test_nc, gold_paras, prompt in all_prompts:\n",
        "for test_nc, gold_paras, prompt in all_prompts[curr_index:]:\n",
        "  # generate number of responses equal to the number of gold paraphrases\n",
        "  time.sleep(15)\n",
        "  num_to_gen = min(len(gold_paras),128)\n",
        "  generated_paras = []\n",
        "  responses = openai.Completion.create(\n",
        "    model=\"text-davinci-002\",\n",
        "    prompt=prompt,\n",
        "    temperature=1,\n",
        "    max_tokens=256,\n",
        "    top_p=1,\n",
        "    frequency_penalty=0,\n",
        "    presence_penalty=0,\n",
        "    n=num_to_gen\n",
        "  )\n",
        "  for r in responses['choices']:\n",
        "    generated_paras.append(r.text.lstrip())\n",
        "\n",
        "  curr_index += 1\n",
        "  print(\"'{}': '{}'\".format(test_nc, generated_paras[0]))\n",
        "  \n",
        "  text_responses[test_nc] = generated_paras"
      ],
      "metadata": {
        "id": "uct4dNiLuhix"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "mydict = {'nc': [], 'paraphrase': []}\n",
        "\n",
        "for nc, paras in text_responses.items():\n",
        "  for p in paras:\n",
        "    mydict['nc'].append(nc)\n",
        "    mydict['paraphrase'].append(p)\n",
        "\n",
        "df = pd.DataFrame(mydict)\n",
        "df"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 424
        },
        "id": "bYok2mSQ5fdf",
        "outputId": "fc6c4158-1ac8-427f-ee93-dc0a9fb07b25"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "                     nc                                         paraphrase\n",
              "0    accounting sleight  a clever or dishonest act or scheme in accounting\n",
              "1    accounting sleight  an accounting method or technique that is inte...\n",
              "2    accounting sleight  a skillful maneuver, especially in financial d...\n",
              "3    accounting sleight                               accounting deception\n",
              "4    accounting sleight  A dishonesty in accounting in order to make fi...\n",
              "..                  ...                                                ...\n",
              "520    wheel efficiency  the efficiency of a wheel in turning or transl...\n",
              "521    wheel efficiency  how much work a wheel can do in relation to th...\n",
              "522    wheel efficiency  The efficiency of a wheel is the Wheel's outpu...\n",
              "523    wheel efficiency  the percentage of power from the engine that i...\n",
              "524    wheel efficiency       how much force is required to move the wheel\n",
              "\n",
              "[525 rows x 2 columns]"
            ],
            "text/html": [
              "\n",
              "  <div id=\"df-a0b1728e-6572-4520-a7ab-77e5704fa437\">\n",
              "    <div 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>nc</th>\n",
              "      <th>paraphrase</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>accounting sleight</td>\n",
              "      <td>a clever or dishonest act or scheme in accounting</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>accounting sleight</td>\n",
              "      <td>an accounting method or technique that is inte...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>accounting sleight</td>\n",
              "      <td>a skillful maneuver, especially in financial d...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>accounting sleight</td>\n",
              "      <td>accounting deception</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>accounting sleight</td>\n",
              "      <td>A dishonesty in accounting in order to make fi...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>...</th>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>520</th>\n",
              "      <td>wheel efficiency</td>\n",
              "      <td>the efficiency of a wheel in turning or transl...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>521</th>\n",
              "      <td>wheel efficiency</td>\n",
              "      <td>how much work a wheel can do in relation to th...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>522</th>\n",
              "      <td>wheel efficiency</td>\n",
              "      <td>The efficiency of a wheel is the Wheel's outpu...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>523</th>\n",
              "      <td>wheel efficiency</td>\n",
              "      <td>the percentage of power from the engine that i...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>524</th>\n",
              "      <td>wheel efficiency</td>\n",
              "      <td>how much force is required to move the wheel</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "<p>525 rows × 2 columns</p>\n",
              "</div>\n",
              "      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-a0b1728e-6572-4520-a7ab-77e5704fa437')\"\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 0 24 24\"\n",
              "       width=\"24px\">\n",
              "    <path d=\"M0 0h24v24H0V0z\" fill=\"none\"/>\n",
              "    <path d=\"M18.56 5.44l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94zm-11 1L8.5 8.5l.94-2.06 2.06-.94-2.06-.94L8.5 2.5l-.94 2.06-2.06.94zm10 10l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94z\"/><path d=\"M17.41 7.96l-1.37-1.37c-.4-.4-.92-.59-1.43-.59-.52 0-1.04.2-1.43.59L10.3 9.45l-7.72 7.72c-.78.78-.78 2.05 0 2.83L4 21.41c.39.39.9.59 1.41.59.51 0 1.02-.2 1.41-.59l7.78-7.78 2.81-2.81c.8-.78.8-2.07 0-2.86zM5.41 20L4 18.59l7.72-7.72 1.47 1.35L5.41 20z\"/>\n",
              "  </svg>\n",
              "      </button>\n",
              "      \n",
              "  <style>\n",
              "    .colab-df-container {\n",
              "      display:flex;\n",
              "      flex-wrap:wrap;\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",
              "    [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-a0b1728e-6572-4520-a7ab-77e5704fa437 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-a0b1728e-6572-4520-a7ab-77e5704fa437');\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",
              "  </div>\n",
              "  "
            ]
          },
          "metadata": {},
          "execution_count": 15
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "df.to_csv('gpt3-responses_new_diff_set.csv')"
      ],
      "metadata": {
        "id": "QUzhuRrzJDop"
      },
      "execution_count": null,
      "outputs": []
    }
  ]
}