Quotation_Generation_Using_Pytroch.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "name": "Quotation_Generation_Using_Pytroch.ipynb",
      "provenance": [],
      "collapsed_sections": [],
      "machine_shape": "hm",
      "include_colab_link": true
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    },
    "accelerator": "GPU",
    "widgets": {
      "application/vnd.jupyter.widget-state+json": {
        "bd34594055094d709dc707946bad5e7f": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "HBoxModel",
          "state": {
            "_view_name": "HBoxView",
            "_dom_classes": [],
            "_model_name": "HBoxModel",
            "_view_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_view_count": null,
            "_view_module_version": "1.5.0",
            "box_style": "",
            "layout": "IPY_MODEL_6d43c9f5602d4ba68a21b2ef74723e96",
            "_model_module": "@jupyter-widgets/controls",
            "children": [
              "IPY_MODEL_d6dd35d2f7684a1ebc8fcbd5f9f6b528",
              "IPY_MODEL_cefa1c09cdd740c3b52ca5808631587d"
            ]
          }
        },
        "6d43c9f5602d4ba68a21b2ef74723e96": {
          "model_module": "@jupyter-widgets/base",
          "model_name": "LayoutModel",
          "state": {
            "_view_name": "LayoutView",
            "grid_template_rows": null,
            "right": null,
            "justify_content": null,
            "_view_module": "@jupyter-widgets/base",
            "overflow": null,
            "_model_module_version": "1.2.0",
            "_view_count": null,
            "flex_flow": null,
            "width": null,
            "min_width": null,
            "border": null,
            "align_items": null,
            "bottom": null,
            "_model_module": "@jupyter-widgets/base",
            "top": null,
            "grid_column": null,
            "overflow_y": null,
            "overflow_x": null,
            "grid_auto_flow": null,
            "grid_area": null,
            "grid_template_columns": null,
            "flex": null,
            "_model_name": "LayoutModel",
            "justify_items": null,
            "grid_row": null,
            "max_height": null,
            "align_content": null,
            "visibility": null,
            "align_self": null,
            "height": null,
            "min_height": null,
            "padding": null,
            "grid_auto_rows": null,
            "grid_gap": null,
            "max_width": null,
            "order": null,
            "_view_module_version": "1.2.0",
            "grid_template_areas": null,
            "object_position": null,
            "object_fit": null,
            "grid_auto_columns": null,
            "margin": null,
            "display": null,
            "left": null
          }
        },
        "d6dd35d2f7684a1ebc8fcbd5f9f6b528": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "IntProgressModel",
          "state": {
            "_view_name": "ProgressView",
            "style": "IPY_MODEL_e6fab950755e424faf0334f87ca3cfa6",
            "_dom_classes": [],
            "description": "",
            "_model_name": "IntProgressModel",
            "bar_style": "success",
            "max": 100,
            "_view_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "value": 100,
            "_view_count": null,
            "_view_module_version": "1.5.0",
            "orientation": "horizontal",
            "min": 0,
            "description_tooltip": null,
            "_model_module": "@jupyter-widgets/controls",
            "layout": "IPY_MODEL_ae2416c135b04c84af1bbc04557e1ea6"
          }
        },
        "cefa1c09cdd740c3b52ca5808631587d": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "HTMLModel",
          "state": {
            "_view_name": "HTMLView",
            "style": "IPY_MODEL_35dffa0ebd824d818a5cd9eb18b11fc0",
            "_dom_classes": [],
            "description": "",
            "_model_name": "HTMLModel",
            "placeholder": "​",
            "_view_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "value": "100% 100/100 [01:10&lt;00:00,  1.44it/s]",
            "_view_count": null,
            "_view_module_version": "1.5.0",
            "description_tooltip": null,
            "_model_module": "@jupyter-widgets/controls",
            "layout": "IPY_MODEL_7d52111aa1b4456d8d5a9a0edf55889e"
          }
        },
        "e6fab950755e424faf0334f87ca3cfa6": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "ProgressStyleModel",
          "state": {
            "_view_name": "StyleView",
            "_model_name": "ProgressStyleModel",
            "description_width": "",
            "_view_module": "@jupyter-widgets/base",
            "_model_module_version": "1.5.0",
            "_view_count": null,
            "_view_module_version": "1.2.0",
            "bar_color": null,
            "_model_module": "@jupyter-widgets/controls"
          }
        },
        "ae2416c135b04c84af1bbc04557e1ea6": {
          "model_module": "@jupyter-widgets/base",
          "model_name": "LayoutModel",
          "state": {
            "_view_name": "LayoutView",
            "grid_template_rows": null,
            "right": null,
            "justify_content": null,
            "_view_module": "@jupyter-widgets/base",
            "overflow": null,
            "_model_module_version": "1.2.0",
            "_view_count": null,
            "flex_flow": null,
            "width": null,
            "min_width": null,
            "border": null,
            "align_items": null,
            "bottom": null,
            "_model_module": "@jupyter-widgets/base",
            "top": null,
            "grid_column": null,
            "overflow_y": null,
            "overflow_x": null,
            "grid_auto_flow": null,
            "grid_area": null,
            "grid_template_columns": null,
            "flex": null,
            "_model_name": "LayoutModel",
            "justify_items": null,
            "grid_row": null,
            "max_height": null,
            "align_content": null,
            "visibility": null,
            "align_self": null,
            "height": null,
            "min_height": null,
            "padding": null,
            "grid_auto_rows": null,
            "grid_gap": null,
            "max_width": null,
            "order": null,
            "_view_module_version": "1.2.0",
            "grid_template_areas": null,
            "object_position": null,
            "object_fit": null,
            "grid_auto_columns": null,
            "margin": null,
            "display": null,
            "left": null
          }
        },
        "35dffa0ebd824d818a5cd9eb18b11fc0": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "DescriptionStyleModel",
          "state": {
            "_view_name": "StyleView",
            "_model_name": "DescriptionStyleModel",
            "description_width": "",
            "_view_module": "@jupyter-widgets/base",
            "_model_module_version": "1.5.0",
            "_view_count": null,
            "_view_module_version": "1.2.0",
            "_model_module": "@jupyter-widgets/controls"
          }
        },
        "7d52111aa1b4456d8d5a9a0edf55889e": {
          "model_module": "@jupyter-widgets/base",
          "model_name": "LayoutModel",
          "state": {
            "_view_name": "LayoutView",
            "grid_template_rows": null,
            "right": null,
            "justify_content": null,
            "_view_module": "@jupyter-widgets/base",
            "overflow": null,
            "_model_module_version": "1.2.0",
            "_view_count": null,
            "flex_flow": null,
            "width": null,
            "min_width": null,
            "border": null,
            "align_items": null,
            "bottom": null,
            "_model_module": "@jupyter-widgets/base",
            "top": null,
            "grid_column": null,
            "overflow_y": null,
            "overflow_x": null,
            "grid_auto_flow": null,
            "grid_area": null,
            "grid_template_columns": null,
            "flex": null,
            "_model_name": "LayoutModel",
            "justify_items": null,
            "grid_row": null,
            "max_height": null,
            "align_content": null,
            "visibility": null,
            "align_self": null,
            "height": null,
            "min_height": null,
            "padding": null,
            "grid_auto_rows": null,
            "grid_gap": null,
            "max_width": null,
            "order": null,
            "_view_module_version": "1.2.0",
            "grid_template_areas": null,
            "object_position": null,
            "object_fit": null,
            "grid_auto_columns": null,
            "margin": null,
            "display": null,
            "left": null
          }
        },
        "f7594daade134d39a5895021def523b3": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "HBoxModel",
          "state": {
            "_view_name": "HBoxView",
            "_dom_classes": [],
            "_model_name": "HBoxModel",
            "_view_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_view_count": null,
            "_view_module_version": "1.5.0",
            "box_style": "",
            "layout": "IPY_MODEL_2bb2d4249aad4aaea8ae198859b9adb0",
            "_model_module": "@jupyter-widgets/controls",
            "children": [
              "IPY_MODEL_a11b435e29b0479e8a6bb259b36dd67b",
              "IPY_MODEL_1408cc473d58441e91eeb427c91eac24"
            ]
          }
        },
        "2bb2d4249aad4aaea8ae198859b9adb0": {
          "model_module": "@jupyter-widgets/base",
          "model_name": "LayoutModel",
          "state": {
            "_view_name": "LayoutView",
            "grid_template_rows": null,
            "right": null,
            "justify_content": null,
            "_view_module": "@jupyter-widgets/base",
            "overflow": null,
            "_model_module_version": "1.2.0",
            "_view_count": null,
            "flex_flow": null,
            "width": null,
            "min_width": null,
            "border": null,
            "align_items": null,
            "bottom": null,
            "_model_module": "@jupyter-widgets/base",
            "top": null,
            "grid_column": null,
            "overflow_y": null,
            "overflow_x": null,
            "grid_auto_flow": null,
            "grid_area": null,
            "grid_template_columns": null,
            "flex": null,
            "_model_name": "LayoutModel",
            "justify_items": null,
            "grid_row": null,
            "max_height": null,
            "align_content": null,
            "visibility": null,
            "align_self": null,
            "height": null,
            "min_height": null,
            "padding": null,
            "grid_auto_rows": null,
            "grid_gap": null,
            "max_width": null,
            "order": null,
            "_view_module_version": "1.2.0",
            "grid_template_areas": null,
            "object_position": null,
            "object_fit": null,
            "grid_auto_columns": null,
            "margin": null,
            "display": null,
            "left": null
          }
        },
        "a11b435e29b0479e8a6bb259b36dd67b": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "IntProgressModel",
          "state": {
            "_view_name": "ProgressView",
            "style": "IPY_MODEL_df8d5fd2ca2c4dde80ec15afdad3f984",
            "_dom_classes": [],
            "description": "",
            "_model_name": "IntProgressModel",
            "bar_style": "success",
            "max": 101,
            "_view_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "value": 101,
            "_view_count": null,
            "_view_module_version": "1.5.0",
            "orientation": "horizontal",
            "min": 0,
            "description_tooltip": null,
            "_model_module": "@jupyter-widgets/controls",
            "layout": "IPY_MODEL_6cf2d224b7bc4f75a424d3da2583dde4"
          }
        },
        "1408cc473d58441e91eeb427c91eac24": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "HTMLModel",
          "state": {
            "_view_name": "HTMLView",
            "style": "IPY_MODEL_2534da0f3d1e4ec69b9944635889c83f",
            "_dom_classes": [],
            "description": "",
            "_model_name": "HTMLModel",
            "placeholder": "​",
            "_view_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "value": "100% 101/101 [05:11&lt;00:00,  3.07s/it]",
            "_view_count": null,
            "_view_module_version": "1.5.0",
            "description_tooltip": null,
            "_model_module": "@jupyter-widgets/controls",
            "layout": "IPY_MODEL_c90fe095f76a40b5958df10fd1fd1ccb"
          }
        },
        "df8d5fd2ca2c4dde80ec15afdad3f984": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "ProgressStyleModel",
          "state": {
            "_view_name": "StyleView",
            "_model_name": "ProgressStyleModel",
            "description_width": "",
            "_view_module": "@jupyter-widgets/base",
            "_model_module_version": "1.5.0",
            "_view_count": null,
            "_view_module_version": "1.2.0",
            "bar_color": null,
            "_model_module": "@jupyter-widgets/controls"
          }
        },
        "6cf2d224b7bc4f75a424d3da2583dde4": {
          "model_module": "@jupyter-widgets/base",
          "model_name": "LayoutModel",
          "state": {
            "_view_name": "LayoutView",
            "grid_template_rows": null,
            "right": null,
            "justify_content": null,
            "_view_module": "@jupyter-widgets/base",
            "overflow": null,
            "_model_module_version": "1.2.0",
            "_view_count": null,
            "flex_flow": null,
            "width": null,
            "min_width": null,
            "border": null,
            "align_items": null,
            "bottom": null,
            "_model_module": "@jupyter-widgets/base",
            "top": null,
            "grid_column": null,
            "overflow_y": null,
            "overflow_x": null,
            "grid_auto_flow": null,
            "grid_area": null,
            "grid_template_columns": null,
            "flex": null,
            "_model_name": "LayoutModel",
            "justify_items": null,
            "grid_row": null,
            "max_height": null,
            "align_content": null,
            "visibility": null,
            "align_self": null,
            "height": null,
            "min_height": null,
            "padding": null,
            "grid_auto_rows": null,
            "grid_gap": null,
            "max_width": null,
            "order": null,
            "_view_module_version": "1.2.0",
            "grid_template_areas": null,
            "object_position": null,
            "object_fit": null,
            "grid_auto_columns": null,
            "margin": null,
            "display": null,
            "left": null
          }
        },
        "2534da0f3d1e4ec69b9944635889c83f": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "DescriptionStyleModel",
          "state": {
            "_view_name": "StyleView",
            "_model_name": "DescriptionStyleModel",
            "description_width": "",
            "_view_module": "@jupyter-widgets/base",
            "_model_module_version": "1.5.0",
            "_view_count": null,
            "_view_module_version": "1.2.0",
            "_model_module": "@jupyter-widgets/controls"
          }
        },
        "c90fe095f76a40b5958df10fd1fd1ccb": {
          "model_module": "@jupyter-widgets/base",
          "model_name": "LayoutModel",
          "state": {
            "_view_name": "LayoutView",
            "grid_template_rows": null,
            "right": null,
            "justify_content": null,
            "_view_module": "@jupyter-widgets/base",
            "overflow": null,
            "_model_module_version": "1.2.0",
            "_view_count": null,
            "flex_flow": null,
            "width": null,
            "min_width": null,
            "border": null,
            "align_items": null,
            "bottom": null,
            "_model_module": "@jupyter-widgets/base",
            "top": null,
            "grid_column": null,
            "overflow_y": null,
            "overflow_x": null,
            "grid_auto_flow": null,
            "grid_area": null,
            "grid_template_columns": null,
            "flex": null,
            "_model_name": "LayoutModel",
            "justify_items": null,
            "grid_row": null,
            "max_height": null,
            "align_content": null,
            "visibility": null,
            "align_self": null,
            "height": null,
            "min_height": null,
            "padding": null,
            "grid_auto_rows": null,
            "grid_gap": null,
            "max_width": null,
            "order": null,
            "_view_module_version": "1.2.0",
            "grid_template_areas": null,
            "object_position": null,
            "object_fit": null,
            "grid_auto_columns": null,
            "margin": null,
            "display": null,
            "left": null
          }
        }
      }
    }
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "view-in-github",
        "colab_type": "text"
      },
      "source": [
        "<a href=\"https://colab.research.google.com/github/santhalakshminarayana/Quotation_Generation_Glove_LSTM_Pytorch/blob/master/Quotation_Generation_Using_Pytroch.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "NJs8GfJIxLUT",
        "colab_type": "code",
        "outputId": "9dbcac4a-91a1-418f-c4ba-d0b542f6e91b",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 122
        }
      },
      "source": [
        "from google.colab import drive\n",
        "drive.mount('/content/drive')"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Go to this URL in a browser: https://accounts.google.com/o/oauth2/auth?client_id=947318989803-6bn6qk8qdgf4n4g3pfee6491hc0brc4i.apps.googleusercontent.com&redirect_uri=urn%3aietf%3awg%3aoauth%3a2.0%3aoob&response_type=code&scope=email%20https%3a%2f%2fwww.googleapis.com%2fauth%2fdocs.test%20https%3a%2f%2fwww.googleapis.com%2fauth%2fdrive%20https%3a%2f%2fwww.googleapis.com%2fauth%2fdrive.photos.readonly%20https%3a%2f%2fwww.googleapis.com%2fauth%2fpeopleapi.readonly\n",
            "\n",
            "Enter your authorization code:\n",
            "··········\n",
            "Mounted at /content/drive\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "TovF7PD7xUub",
        "colab_type": "code",
        "outputId": "e8ee8d41-efec-4903-ddc8-16b305a15d03",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "source": [
        "% cd /content/drive/My\\ Drive/Colab\\ Notebooks/Quotation_Generation/"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "/content/drive/My Drive/Colab Notebooks/Quotation_Generation\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "UMLKRdJkLtI2",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "from collections import  defaultdict\n",
        "\n",
        "import numpy as np\n",
        "import seaborn as sns \n",
        "import matplotlib.pyplot as plt\n",
        "import math\n",
        "sns.set(palette = 'gist_earth', style='white')\n",
        "\n",
        "import torch\n",
        "import torch.nn as nn\n",
        "import torch.optim as optim\n",
        "import torch.nn.functional as F\n",
        "\n",
        "from tqdm import tqdm_notebook as tqdm\n",
        "\n",
        "from sklearn.manifold import TSNE"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "TLp5gG6PyQ9e",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "quotes = []\n",
        "max_len = 0\n",
        "min_len = 5\n",
        "sent_len_dic = defaultdict(int)\n",
        "with open('quotes.txt', 'r') as f:\n",
        "  while True:\n",
        "    quote = f.readline()\n",
        "    if not quote:\n",
        "      break\n",
        "    words = quote.split(' ')\n",
        "    sent_len = len(words)\n",
        "    if sent_len < min_len:\n",
        "      continue\n",
        "    max_len = sent_len if sent_len > max_len else max_len\n",
        "    sent_len_dic[sent_len] += 1\n",
        "    flag = False\n",
        "    for i in range(0,10):\n",
        "      if str(i) in quote:\n",
        "        flag = True\n",
        "        break\n",
        "    if flag == True:\n",
        "      continue\n",
        "    quote = quote.replace(',', ' ,').replace('.', ' .').replace('_','').strip('\\n') + ' ;'\n",
        "    quotes.append(quote.lower())"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "QLyfEDHUjxDd",
        "colab_type": "code",
        "outputId": "0c84b06c-d534-4842-b56b-07560509d0d5",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 68
        }
      },
      "source": [
        "print(f'Sentences with\\nmin count words : {min_len}\\nmax count words : {max_len}')"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Sentences with\n",
            "min count words : 5\n",
            "max count words : 111\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "EsDrBGoWygRT",
        "colab_type": "code",
        "outputId": "8a4f8e62-59ee-48a1-d3c9-78e3736f74ab",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "source": [
        "len(quotes)"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "3831"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 6
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "XCI0JFJPy4zO",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "words_dict = dict()\n",
        "def unique_words(data_list):\n",
        "  for x in data_list:\n",
        "    for word in x.split(' '):\n",
        "      words_dict[word] = 1\n",
        "  return words_dict.keys()"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "iJuJEO1U0wlo",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "words = list(unique_words(quotes))\n",
        "words.sort()"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "QDqZwqq-5D7Y",
        "colab_type": "code",
        "outputId": "e8300783-fcb2-4e45-ac06-20711f1670db",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "source": [
        "len(words)"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "6361"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 9
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "HLkaeFAI8x7Q",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "word_to_int = dict()\n",
        "int_to_word = dict()\n",
        "vocab_len = len(words)\n",
        "for i in range(0, vocab_len):\n",
        "  word_to_int[words[i]] = i \n",
        "  int_to_word[i] = words[i]"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "CeEnw4u7bsxW",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "co_occ_matrix = defaultdict(int)\n",
        "window = 5 # not greater than 5\n",
        "for sent in quotes:\n",
        "  words = sent.split(' ')\n",
        "  # first window \n",
        "  for i in range(0, window):\n",
        "    for j in range(i+1, window):\n",
        "      weight = 1/(j-i)\n",
        "      ind_1 = word_to_int[words[i]]\n",
        "      ind_2 = word_to_int[words[j]]\n",
        "      if ind_1 == ind_2:\n",
        "        continue\n",
        "      co_occ_matrix[(ind_1, ind_2)] += weight\n",
        "      co_occ_matrix[(ind_2, ind_1)] += weight\n",
        "  # rest windows\n",
        "  for i in range(window, len(words)):\n",
        "    for j in range(i-window, i):\n",
        "      weight = 1/(i-j)\n",
        "      ind_1 = word_to_int[words[i]]\n",
        "      ind_2 = word_to_int[words[j]]\n",
        "      if ind_1 == ind_2:\n",
        "        continue\n",
        "      co_occ_matrix[(ind_1, ind_2)] += weight\n",
        "      co_occ_matrix[(ind_2, ind_1)] += weight"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "1csW6dDKMoj1",
        "colab_type": "code",
        "outputId": "ad32335e-92e0-4eab-e0aa-8ad1df912b60",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 285
        }
      },
      "source": [
        "topk, ind = torch.topk(torch.tensor(list(co_occ_matrix.values())), 200)\n",
        "sns.distplot(topk)"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<matplotlib.axes._subplots.AxesSubplot at 0x7ff1f3f938d0>"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 12
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYkAAAD7CAYAAACfQGjDAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n8oohQUREXjEkiIjIq/8PXNWYHA/DK20AAAAASUVORK5CYII=\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "J_FCAJVIL-1F",
        "colab_type": "code",
        "outputId": "e8800828-b823-4ce4-c80d-b0293695b462",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 285
        }
      },
      "source": [
        "elem_den = []\n",
        "for i in co_occ_matrix.values():\n",
        "  if i > 0 and i < 10:\n",
        "    elem_den.append(i)\n",
        "sns.violinplot(x = elem_den)"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<matplotlib.axes._subplots.AxesSubplot at 0x7ff1f1c90f98>"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 13
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAV0AAAD7CAYAAADJukfwAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "sKEmnNzjeHRR",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "tot_pairs = len(co_occ_matrix)\n",
        "p1, p2,occs = list(), list(), list()\n",
        "for i in co_occ_matrix.keys():\n",
        "    p1.append(i[0])\n",
        "    p2.append(i[1])\n",
        "    occs.append(co_occ_matrix[i])\n",
        "p1 = np.array(p1)\n",
        "p2 = np.array(p2)\n",
        "occs = np.array(occs)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "tmKvsPupQ-Z6",
        "colab_type": "code",
        "outputId": "5fbf38dd-a640-4ff8-f944-1398f3d5637d",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "source": [
        "occs.size"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "249808"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 15
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Ontxjlf0cBjm",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "def get_batch(batch_size):\n",
        "  ind = np.random.permutation(occs.size).tolist()\n",
        "  i = 0\n",
        "  for i in range(0, tot_pairs, batch_size):\n",
        "    batch_ids = ind[i:i+batch_size]\n",
        "    yield p1[batch_ids], p2[batch_ids], occs[batch_ids]"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "lF4e-6wkXcoI",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "device = None\n",
        "if torch.cuda.is_available():\n",
        "    device = torch.device(\"cuda:0\")\n",
        "else:\n",
        "    device = torch.device(\"cpu\")"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "7VOYaxcZVRwD",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "class Glove(nn.Module) :\n",
        "  def __init__(self, vocab_len, num_dim,):\n",
        "    super(Glove, self).__init__()\n",
        "    self.ui = nn.Embedding(vocab_len, num_dim).to(device)\n",
        "    self.uj = nn.Embedding(vocab_len, num_dim).to(device)\n",
        "    self.bi = nn.Embedding(vocab_len, 1).to(device)\n",
        "    self.bj = nn.Embedding(vocab_len, 1).to(device)\n",
        "\n",
        "    self.ui.weight.data.uniform_(-1,1)\n",
        "    self.uj.weight.data.uniform_(-1,1)\n",
        "    self.bi.weight.data.zero_()\n",
        "    self.bj.weight.data.zero_()\n",
        "  \n",
        "  def forward(self, i_vecs, j_vecs):\n",
        "    i_s = self.ui(i_vecs)\n",
        "    j_s = self.uj(j_vecs)\n",
        "    b_i = torch.squeeze(self.bi(i_vecs))\n",
        "    b_j = torch.squeeze(self.bj(j_vecs))\n",
        "    return torch.sum(i_s * j_s, dim = 1) + b_i + b_j"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "GfX16YdopLAJ",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "num_dim = 128\n",
        "glove = Glove(vocab_len, num_dim).to(device)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "0wuj2wyAI1C9",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "def f_x(x, x_max, alpha):\n",
        "  x = (x/x_max)**alpha\n",
        "  return torch.min(x, torch.ones_like(x)).to(device)\n",
        "\n",
        "def weight_mse(w_x, x, log_x):\n",
        "  loss = w_x * F.mse_loss(x, log_x, reduction='none')\n",
        "  return torch.mean(loss).to(device)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "yzCMv1ubl7NO",
        "colab_type": "code",
        "outputId": "7dae85eb-11d5-4f35-eaa5-e9028f04a381",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1000,
          "referenced_widgets": [
            "bd34594055094d709dc707946bad5e7f",
            "6d43c9f5602d4ba68a21b2ef74723e96",
            "d6dd35d2f7684a1ebc8fcbd5f9f6b528",
            "cefa1c09cdd740c3b52ca5808631587d",
            "e6fab950755e424faf0334f87ca3cfa6",
            "ae2416c135b04c84af1bbc04557e1ea6",
            "35dffa0ebd824d818a5cd9eb18b11fc0",
            "7d52111aa1b4456d8d5a9a0edf55889e"
          ]
        }
      },
      "source": [
        "epochs = 100\n",
        "batch_size = 512\n",
        "batch_count = 0\n",
        "x_max = 1\n",
        "alpha = 0.75\n",
        "loss_trace = []\n",
        "n_bathces = int(math.floor(len(occs)/batch_size))\n",
        "optimizer = optim.Adagrad(glove.parameters(), lr=0.05)\n",
        "for epoch in tqdm(range(epochs)):\n",
        "  batch_count = -1\n",
        "  for ind_1, ind_2, occ in get_batch(batch_size):\n",
        "    batch_count += 1\n",
        "    occ = torch.FloatTensor(occ).to(device)\n",
        "    optimizer.zero_grad()\n",
        "    y_hat = glove.forward(torch.tensor(ind_1).to(device), \n",
        "                          torch.tensor(ind_2).to(device))\n",
        "    w_x = f_x(occ, x_max, alpha)\n",
        "    loss = weight_mse(w_x, y_hat, torch.log(occ))\n",
        "    loss.backward()\n",
        "    optimizer.step()\n",
        "    loss_trace.append(loss.item())\n",
        "    if batch_count == n_bathces:\n",
        "      break\n",
        "  print(f\"Epoch : {epoch + 1} ----> Loss : {loss_trace[-1]}\")"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "application/vnd.jupyter.widget-view+json": {
              "model_id": "bd34594055094d709dc707946bad5e7f",
              "version_minor": 0,
              "version_major": 2
            },
            "text/plain": [
              "HBox(children=(IntProgress(value=0), HTML(value='')))"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "stream",
          "text": [
            "Epoch : 1 ----> Loss : 4.796401500701904\n",
            "Epoch : 2 ----> Loss : 2.014411211013794\n",
            "Epoch : 3 ----> Loss : 1.426303744316101\n",
            "Epoch : 4 ----> Loss : 1.0219517946243286\n",
            "Epoch : 5 ----> Loss : 0.9586169719696045\n",
            "Epoch : 6 ----> Loss : 0.7812556624412537\n",
            "Epoch : 7 ----> Loss : 0.7010027170181274\n",
            "Epoch : 8 ----> Loss : 0.6730878353118896\n",
            "Epoch : 9 ----> Loss : 0.505446195602417\n",
            "Epoch : 10 ----> Loss : 0.48864707350730896\n",
            "Epoch : 11 ----> Loss : 0.42977482080459595\n",
            "Epoch : 12 ----> Loss : 0.4258069694042206\n",
            "Epoch : 13 ----> Loss : 0.40871748328208923\n",
            "Epoch : 14 ----> Loss : 0.3873564898967743\n",
            "Epoch : 15 ----> Loss : 0.3715595602989197\n",
            "Epoch : 16 ----> Loss : 0.38262832164764404\n",
            "Epoch : 17 ----> Loss : 0.3344598412513733\n",
            "Epoch : 18 ----> Loss : 0.3035099506378174\n",
            "Epoch : 19 ----> Loss : 0.29445740580558777\n",
            "Epoch : 20 ----> Loss : 0.2994334101676941\n",
            "Epoch : 21 ----> Loss : 0.20549406111240387\n",
            "Epoch : 22 ----> Loss : 0.24363911151885986\n",
            "Epoch : 23 ----> Loss : 0.2681289613246918\n",
            "Epoch : 24 ----> Loss : 0.24000214040279388\n",
            "Epoch : 25 ----> Loss : 0.2526589334011078\n",
            "Epoch : 26 ----> Loss : 0.2096152901649475\n",
            "Epoch : 27 ----> Loss : 0.20147758722305298\n",
            "Epoch : 28 ----> Loss : 0.18120785057544708\n",
            "Epoch : 29 ----> Loss : 0.19716937839984894\n",
            "Epoch : 30 ----> Loss : 0.17901384830474854\n",
            "Epoch : 31 ----> Loss : 0.1933664083480835\n",
            "Epoch : 32 ----> Loss : 0.16095784306526184\n",
            "Epoch : 33 ----> Loss : 0.1736927330493927\n",
            "Epoch : 34 ----> Loss : 0.1567644625902176\n",
            "Epoch : 35 ----> Loss : 0.19178061187267303\n",
            "Epoch : 36 ----> Loss : 0.1426701694726944\n",
            "Epoch : 37 ----> Loss : 0.17866656184196472\n",
            "Epoch : 38 ----> Loss : 0.15436622500419617\n",
            "Epoch : 39 ----> Loss : 0.18198798596858978\n",
            "Epoch : 40 ----> Loss : 0.16307803988456726\n",
            "Epoch : 41 ----> Loss : 0.16314055025577545\n",
            "Epoch : 42 ----> Loss : 0.15512752532958984\n",
            "Epoch : 43 ----> Loss : 0.1407899260520935\n",
            "Epoch : 44 ----> Loss : 0.16897889971733093\n",
            "Epoch : 45 ----> Loss : 0.14180248975753784\n",
            "Epoch : 46 ----> Loss : 0.15545207262039185\n",
            "Epoch : 47 ----> Loss : 0.15387989580631256\n",
            "Epoch : 48 ----> Loss : 0.12659546732902527\n",
            "Epoch : 49 ----> Loss : 0.15585926175117493\n",
            "Epoch : 50 ----> Loss : 0.13493146002292633\n",
            "Epoch : 51 ----> Loss : 0.12801656126976013\n",
            "Epoch : 52 ----> Loss : 0.13777688145637512\n",
            "Epoch : 53 ----> Loss : 0.09616035968065262\n",
            "Epoch : 54 ----> Loss : 0.14205625653266907\n",
            "Epoch : 55 ----> Loss : 0.11623765528202057\n",
            "Epoch : 56 ----> Loss : 0.12292160093784332\n",
            "Epoch : 57 ----> Loss : 0.1190635934472084\n",
            "Epoch : 58 ----> Loss : 0.11154908686876297\n",
            "Epoch : 59 ----> Loss : 0.12162309885025024\n",
            "Epoch : 60 ----> Loss : 0.1064210757613182\n",
            "Epoch : 61 ----> Loss : 0.10693100094795227\n",
            "Epoch : 62 ----> Loss : 0.11972589045763016\n",
            "Epoch : 63 ----> Loss : 0.11175313591957092\n",
            "Epoch : 64 ----> Loss : 0.09873320907354355\n",
            "Epoch : 65 ----> Loss : 0.10549955815076828\n",
            "Epoch : 66 ----> Loss : 0.1005711704492569\n",
            "Epoch : 67 ----> Loss : 0.11455339938402176\n",
            "Epoch : 68 ----> Loss : 0.09063642472028732\n",
            "Epoch : 69 ----> Loss : 0.10180366784334183\n",
            "Epoch : 70 ----> Loss : 0.14373937249183655\n",
            "Epoch : 71 ----> Loss : 0.11405614763498306\n",
            "Epoch : 72 ----> Loss : 0.09900388866662979\n",
            "Epoch : 73 ----> Loss : 0.10204146057367325\n",
            "Epoch : 74 ----> Loss : 0.11652762442827225\n",
            "Epoch : 75 ----> Loss : 0.1080511212348938\n",
            "Epoch : 76 ----> Loss : 0.10836173593997955\n",
            "Epoch : 77 ----> Loss : 0.0980234146118164\n",
            "Epoch : 78 ----> Loss : 0.10077328979969025\n",
            "Epoch : 79 ----> Loss : 0.10664699971675873\n",
            "Epoch : 80 ----> Loss : 0.08906294405460358\n",
            "Epoch : 81 ----> Loss : 0.0909762755036354\n",
            "Epoch : 82 ----> Loss : 0.0957847461104393\n",
            "Epoch : 83 ----> Loss : 0.07920680940151215\n",
            "Epoch : 84 ----> Loss : 0.08984296023845673\n",
            "Epoch : 85 ----> Loss : 0.09093265980482101\n",
            "Epoch : 86 ----> Loss : 0.089451864361763\n",
            "Epoch : 87 ----> Loss : 0.08259624987840652\n",
            "Epoch : 88 ----> Loss : 0.08016000688076019\n",
            "Epoch : 89 ----> Loss : 0.08158089220523834\n",
            "Epoch : 90 ----> Loss : 0.09190693497657776\n",
            "Epoch : 91 ----> Loss : 0.10357186198234558\n",
            "Epoch : 92 ----> Loss : 0.09238259494304657\n",
            "Epoch : 93 ----> Loss : 0.07649978250265121\n",
            "Epoch : 94 ----> Loss : 0.07427524775266647\n",
            "Epoch : 95 ----> Loss : 0.07595456391572952\n",
            "Epoch : 96 ----> Loss : 0.07658737897872925\n",
            "Epoch : 97 ----> Loss : 0.08572196960449219\n",
            "Epoch : 98 ----> Loss : 0.08962361514568329\n",
            "Epoch : 99 ----> Loss : 0.07171893864870071\n",
            "Epoch : 100 ----> Loss : 0.06875928491353989\n",
            "\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "iZ5NxoeLkdT-",
        "colab_type": "code",
        "outputId": "a20442c9-de3b-4179-9a81-c6e039c0ebe2",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 285
        }
      },
      "source": [
        "plt.plot(loss_trace)"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "[<matplotlib.lines.Line2D at 0x7ff1e14b8e48>]"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 22
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAX4AAAD7CAYAAABt0P8jAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nDwAOQ/EDgMNQ/ADgMBQ/ADgMxQ8ADvN/E23/bcLFd90AAAAASUVORK5CYII=\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "fv7_ZPOwmm6a",
        "colab_type": "code",
        "outputId": "89ebb2da-104c-424b-9396-4a26c975c7b8",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 85
        }
      },
      "source": [
        "torch.save(glove, 'glove.pth')"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.6/dist-packages/torch/serialization.py:292: UserWarning: Couldn't retrieve source code for container of type Glove. It won't be checked for correctness upon loading.\n",
            "  \"type \" + obj.__name__ + \". It won't be checked \"\n",
            "/usr/local/lib/python3.6/dist-packages/torch/serialization.py:292: UserWarning: Couldn't retrieve source code for container of type Embedding. It won't be checked for correctness upon loading.\n",
            "  \"type \" + obj.__name__ + \". It won't be checked \"\n"
          ],
          "name": "stderr"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "SjRblVKbqqSn",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "glove = torch.load('glove.pth',device)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "ErlX-NAkrhwm",
        "colab_type": "code",
        "outputId": "a82ec64c-1bcc-4656-9040-915c1cee82c8",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 119
        }
      },
      "source": [
        "glove.eval()"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "Glove(\n",
              "  (ui): Embedding(6361, 128)\n",
              "  (uj): Embedding(6361, 128)\n",
              "  (bi): Embedding(6361, 1)\n",
              "  (bj): Embedding(6361, 1)\n",
              ")"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 15
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "jnmwIedtfY1k",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "emb_i = glove.ui.weight.cpu().data.numpy()\n",
        "emb_j = glove.uj.weight.cpu().data.numpy()\n",
        "emb = emb_i + emb_j"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "SYXrZ0TGrmsH",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "tsne = TSNE(n_components=2, perplexity=40, n_iter=300,metric='cosine')\n",
        "n_points = 100\n",
        "emb_tsne = tsne.fit_transform(emb[:n_points, :])"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "LKF7Jx04s4kX",
        "colab_type": "code",
        "outputId": "dacd85f2-0fd8-4687-c555-c7ee36c88e82",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 920
        }
      },
      "source": [
        "labels = list(word_to_int.keys())[:n_points]\n",
        "x = emb_tsne[:,0]\n",
        "y = emb_tsne[:,1]\n",
        "plt.figure(figsize=(16, 16)) \n",
        "for i in range(n_points):\n",
        "  plt.scatter(x[i],y[i])\n",
        "  plt.annotate(labels[i], xy=(x[i], y[i]), xytext=(5, 2),\n",
        "               textcoords='offset points', ha='right', va='bottom')"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAA6QAAAOHCAYAAAAwh45FAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nAAAAihCkAAAAFPE/eUP28VHloBIAAAAASUVORK5CYII=\n",
            "text/plain": [
              "<Figure size 1152x1152 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "R-htHDXghDFq",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "max_sent_len = 50\n",
        "window = 5\n",
        "sequences, next_words = [], []\n",
        "for quote in quotes:\n",
        "  words = quote.split(' ')\n",
        "  for i in range(0,len(words) - window + 1):\n",
        "    sequences.append(words[i:i+window])\n",
        "    if (i + window) < len(words):\n",
        "      next_words.append(words[i+window])\n",
        "    else:\n",
        "      next_words.append(';')"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "7pQ7IA7PnncW",
        "colab_type": "code",
        "outputId": "5ad1730a-b823-4132-d4b0-5dd32e716c66",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 391
        }
      },
      "source": [
        "print(\"{0:^40} {1}\".format('Sequence', 'Next_Word'))\n",
        "print('-'*50)\n",
        "for i in range(20):\n",
        "  print(f\"{' '.join(sequences[i]) : <40} {next_words[i] : ^10}\")"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "                Sequence                 Next_Word\n",
            "--------------------------------------------------\n",
            "acting is the least mysterious               of    \n",
            "is the least mysterious of                  all    \n",
            "the least mysterious of all                crafts  \n",
            "least mysterious of all crafts               .     \n",
            "mysterious of all crafts .                whenever \n",
            "of all crafts . whenever                     we    \n",
            "all crafts . whenever we                    want   \n",
            "crafts . whenever we want                something \n",
            ". whenever we want something                from   \n",
            "whenever we want something from           somebody \n",
            "we want something from somebody              or    \n",
            "want something from somebody or             when   \n",
            "something from somebody or when              we    \n",
            "from somebody or when we                    want   \n",
            "somebody or when we want                     to    \n",
            "or when we want to                          hide   \n",
            "when we want to hide                     something \n",
            "we want to hide something                    or    \n",
            "want to hide something or                 pretend  \n",
            "to hide something or pretend                 ,     \n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "NFB6axzVn2rc",
        "colab_type": "code",
        "outputId": "540a2459-6695-4161-8949-30c08ded6a4d",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "source": [
        "tot_seq = len(next_words)\n",
        "print(tot_seq)"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "65915\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "J2r9BKCFtQt5",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "X = np.zeros((tot_seq, window, num_dim))\n",
        "Y = np.zeros((tot_seq,vocab_len))\n",
        "for i,seq in enumerate(sequences):\n",
        "  for j,word in enumerate(seq):\n",
        "    num_id = word_to_int[word]\n",
        "    X[i][j] = np.squeeze(emb[num_id])\n",
        "  num_id = word_to_int[next_words[i]]\n",
        "  Y[i][num_id] = 1"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "89H1CV2fuOgB",
        "colab_type": "code",
        "outputId": "738fbfb6-e84a-4231-8140-e60e39e57b52",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "source": [
        "X.shape, Y.shape"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "((65915, 5, 128), (65915, 6361))"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 21
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "wdy4YZYt3OOE",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "def get_batches_x(tot_seq, batch_size):\n",
        "  ind = np.random.permutation(tot_seq).tolist()\n",
        "  i = 0\n",
        "  for i in range(0, tot_seq, batch_size):\n",
        "    batch_ids = ind[i:i+batch_size]\n",
        "    yield X[batch_ids], Y[batch_ids]"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "JfUXtQsQQl8f",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "class Quote_Generator(nn.Module):\n",
        "  def __init__(self, embed_size, hidden_size, vocab_len):\n",
        "    super(Quote_Generator, self).__init__()\n",
        "    self.hidden_size = hidden_size\n",
        "    self.lstm = nn.LSTM(embed_size, hidden_size, batch_first = True).to(device)\n",
        "    self.dropout = nn.Dropout(0.4)\n",
        "    self.dense = nn.Linear(hidden_size*5, vocab_len).to(device)\n",
        "  \n",
        "  def forward(self, x, prev_state):\n",
        "    output, state = self.lstm(x)\n",
        "    output = self.dropout(output)\n",
        "    logits = self.dense(output.reshape(-1, hidden_size*5))\n",
        "    return logits, state\n",
        "\n",
        "  def zero_states(self, batch_size):\n",
        "    return (torch.zeros(1, batch_size, self.hidden_size).to(device), \n",
        "            torch.zeros(1, batch_size, self.hidden_size).to(device))"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "i8b9nmmsN_DS",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "def entropy_loss(y, y_hat):\n",
        "  y_hat = F.softmax(y_hat, dim = 1)\n",
        "  ll = - (y * torch.log(y_hat))\n",
        "  return torch.sum(ll, dim = 1).mean().to(device)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "vvDBWuQm1PaZ",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "embed_size = 128\n",
        "hidden_size = 64\n",
        "qt_gen = Quote_Generator(embed_size, hidden_size, vocab_len).to(device)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "8hsAzhL912FI",
        "colab_type": "code",
        "outputId": "76fc199b-76e9-4620-e4d2-a239306ae765",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 236,
          "referenced_widgets": [
            "f7594daade134d39a5895021def523b3",
            "2bb2d4249aad4aaea8ae198859b9adb0",
            "a11b435e29b0479e8a6bb259b36dd67b",
            "1408cc473d58441e91eeb427c91eac24",
            "df8d5fd2ca2c4dde80ec15afdad3f984",
            "6cf2d224b7bc4f75a424d3da2583dde4",
            "2534da0f3d1e4ec69b9944635889c83f",
            "c90fe095f76a40b5958df10fd1fd1ccb"
          ]
        }
      },
      "source": [
        "epochs = 101\n",
        "batch_size = 4096\n",
        "losses = []\n",
        "optimizer = torch.optim.Adam(qt_gen.parameters(), lr=0.001)\n",
        "for epoch in tqdm(range(epochs)):\n",
        "  batches = get_batches_x(tot_seq, batch_size)\n",
        "  h_h, h_c = qt_gen.zero_states(batch_size)\n",
        "  for x,y in batches:\n",
        "    qt_gen.train()\n",
        "    optimizer.zero_grad()\n",
        "    x = torch.tensor(x).float().to(device)\n",
        "    y = torch.tensor(y).long().to(device)\n",
        "    logits, (h_h, h_c) = qt_gen(x, (h_h, h_c))\n",
        "    loss = entropy_loss(y, logits)\n",
        "    h_h.detach()\n",
        "    h_c.detach()\n",
        "    loss.backward()\n",
        "    _ = nn.utils.clip_grad_norm_(qt_gen.parameters(), 5)\n",
        "    optimizer.step()\n",
        "    losses.append(loss.item())\n",
        "  if (epoch) % 10 == 0:\n",
        "    print(f\"Epoch : {epoch} ----> Loss : {np.array(losses).mean()}\")\n",
        "  losses = []"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "application/vnd.jupyter.widget-view+json": {
              "model_id": "f7594daade134d39a5895021def523b3",
              "version_minor": 0,
              "version_major": 2
            },
            "text/plain": [
              "HBox(children=(IntProgress(value=0, max=101), HTML(value='')))"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "stream",
          "text": [
            "Epoch : 0 ----> Loss : 1.981710391886094\n",
            "Epoch : 10 ----> Loss : 1.9763988116208244\n",
            "Epoch : 20 ----> Loss : 1.9651567234712488\n",
            "Epoch : 30 ----> Loss : 1.9701273721807144\n",
            "Epoch : 40 ----> Loss : 1.9645351802601534\n",
            "Epoch : 50 ----> Loss : 1.9601879961350386\n",
            "Epoch : 60 ----> Loss : 1.9378872478709501\n",
            "Epoch : 70 ----> Loss : 1.958003415780909\n",
            "Epoch : 80 ----> Loss : 1.9386976045720719\n",
            "Epoch : 90 ----> Loss : 1.9374018486808329\n",
            "Epoch : 100 ----> Loss : 1.945599780363195\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "ufyG0ZNPDI8V",
        "colab_type": "code",
        "outputId": "f8558b8c-5681-4a80-8aeb-54d3399c5bd0",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 153
        }
      },
      "source": [
        "torch.save(qt_gen, 'qt_gen.pth')"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.6/dist-packages/torch/serialization.py:292: UserWarning: Couldn't retrieve source code for container of type Quote_Generator. It won't be checked for correctness upon loading.\n",
            "  \"type \" + obj.__name__ + \". It won't be checked \"\n",
            "/usr/local/lib/python3.6/dist-packages/torch/serialization.py:292: UserWarning: Couldn't retrieve source code for container of type LSTM. It won't be checked for correctness upon loading.\n",
            "  \"type \" + obj.__name__ + \". It won't be checked \"\n",
            "/usr/local/lib/python3.6/dist-packages/torch/serialization.py:292: UserWarning: Couldn't retrieve source code for container of type Dropout. It won't be checked for correctness upon loading.\n",
            "  \"type \" + obj.__name__ + \". It won't be checked \"\n",
            "/usr/local/lib/python3.6/dist-packages/torch/serialization.py:292: UserWarning: Couldn't retrieve source code for container of type Linear. It won't be checked for correctness upon loading.\n",
            "  \"type \" + obj.__name__ + \". It won't be checked \"\n"
          ],
          "name": "stderr"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "yGjqOjixDTha",
        "colab_type": "code",
        "outputId": "3c9f8bfa-e359-4d5d-d3c1-dde174a1a3ce",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 102
        }
      },
      "source": [
        "qt_gen = torch.load('qt_gen.pth', device)\n",
        "qt_gen.eval()"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "Quote_Generator(\n",
              "  (lstm): LSTM(128, 64, batch_first=True)\n",
              "  (dropout): Dropout(p=0.4, inplace=False)\n",
              "  (dense): Linear(in_features=320, out_features=6361, bias=True)\n",
              ")"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 40
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "1qixxzopJO0D",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "def predict(x, temp):\n",
        "  probs = F.softmax(x / temp, dim = 0)\n",
        "  probs = np.squeeze(probs.detach().cpu().numpy())\n",
        "  ind = np.random.choice(6361, 1, p = probs)\n",
        "  return ind[0]"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "FEqlQATKGO19",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "generated_text = ['there','is','no','one','love']\n",
        "curr_len = 0\n",
        "embeds = []\n",
        "is_end = word_to_int[';']\n",
        "\n",
        "qt_gen.eval()\n",
        "\n",
        "for i in generated_text:\n",
        "  embeds.append(emb[word_to_int[i]])\n",
        "\n",
        "while(curr_len < 50):\n",
        "  curr_len += 1\n",
        "  input_tensor = torch.Tensor(embeds).view(1,5,128).float().to(device)\n",
        "  h_h, h_c = qt_gen.zero_states(1)\n",
        "  output, (h_h, h_c) = qt_gen(input_tensor, (h_h,h_c))\n",
        "  word_ind = predict(output[-1], 1.6)\n",
        "  embeds[0][:4].tolist().extend(emb[word_ind])\n",
        "  generated_text.append(int_to_word[word_ind])\n",
        "  if word_ind == is_end:\n",
        "    break"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "vp0OgyhHGOtx",
        "colab_type": "code",
        "outputId": "c62bf342-d4d1-42a1-c3c7-84752374a995",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 54
        }
      },
      "source": [
        "' '.join(generated_text)"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "\"there is no one love or all just or all for the we be so no are , gone in . never hope too that's hope we really in his , gets life your because . of too you the them all yourself . that often all mother everything no in you pain , must\""
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 63
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "GYtRbd832Hyw",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "torch.cuda.empty_cache()"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "VJQbodGICwRP",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        ""
      ],
      "execution_count": 0,
      "outputs": []
    }
  ]
}