🦜🔗_Chat_with_PDFs_Custom_Knowledge_ChatGPT_with_LangChain.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "provenance": [],
      "include_colab_link": true
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    },
    "language_info": {
      "name": "python"
    },
    "widgets": {
      "application/vnd.jupyter.widget-state+json": {
        "bf4eac20add04b57a1af538f80f2980c": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "TextModel",
          "model_module_version": "1.5.0",
          "state": {
            "_dom_classes": [],
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "TextModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/controls",
            "_view_module_version": "1.5.0",
            "_view_name": "TextView",
            "continuous_update": true,
            "description": "",
            "description_tooltip": null,
            "disabled": false,
            "layout": "IPY_MODEL_867b911c7aac4dc7bdf6923d3a464c1e",
            "placeholder": "Please enter your question:",
            "style": "IPY_MODEL_a32ffe2bd6974b6eaa8a3081a8c1816b",
            "value": ""
          }
        },
        "867b911c7aac4dc7bdf6923d3a464c1e": {
          "model_module": "@jupyter-widgets/base",
          "model_name": "LayoutModel",
          "model_module_version": "1.2.0",
          "state": {
            "_model_module": "@jupyter-widgets/base",
            "_model_module_version": "1.2.0",
            "_model_name": "LayoutModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "LayoutView",
            "align_content": null,
            "align_items": null,
            "align_self": null,
            "border": null,
            "bottom": null,
            "display": null,
            "flex": null,
            "flex_flow": null,
            "grid_area": null,
            "grid_auto_columns": null,
            "grid_auto_flow": null,
            "grid_auto_rows": null,
            "grid_column": null,
            "grid_gap": null,
            "grid_row": null,
            "grid_template_areas": null,
            "grid_template_columns": null,
            "grid_template_rows": null,
            "height": null,
            "justify_content": null,
            "justify_items": null,
            "left": null,
            "margin": null,
            "max_height": null,
            "max_width": null,
            "min_height": null,
            "min_width": null,
            "object_fit": null,
            "object_position": null,
            "order": null,
            "overflow": null,
            "overflow_x": null,
            "overflow_y": null,
            "padding": null,
            "right": null,
            "top": null,
            "visibility": null,
            "width": null
          }
        },
        "a32ffe2bd6974b6eaa8a3081a8c1816b": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "DescriptionStyleModel",
          "model_module_version": "1.5.0",
          "state": {
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "DescriptionStyleModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "StyleView",
            "description_width": ""
          }
        },
        "7a260cc663614805b499690a5a2cb5be": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "HTMLModel",
          "model_module_version": "1.5.0",
          "state": {
            "_dom_classes": [],
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "HTMLModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/controls",
            "_view_module_version": "1.5.0",
            "_view_name": "HTMLView",
            "description": "",
            "description_tooltip": null,
            "layout": "IPY_MODEL_93571528b372443592d502bebd336ca0",
            "placeholder": "​",
            "style": "IPY_MODEL_94ce74af3988415ca43a5d23accce7ca",
            "value": "<b>User:</b> Who created transformers?"
          }
        },
        "93571528b372443592d502bebd336ca0": {
          "model_module": "@jupyter-widgets/base",
          "model_name": "LayoutModel",
          "model_module_version": "1.2.0",
          "state": {
            "_model_module": "@jupyter-widgets/base",
            "_model_module_version": "1.2.0",
            "_model_name": "LayoutModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "LayoutView",
            "align_content": null,
            "align_items": null,
            "align_self": null,
            "border": null,
            "bottom": null,
            "display": null,
            "flex": null,
            "flex_flow": null,
            "grid_area": null,
            "grid_auto_columns": null,
            "grid_auto_flow": null,
            "grid_auto_rows": null,
            "grid_column": null,
            "grid_gap": null,
            "grid_row": null,
            "grid_template_areas": null,
            "grid_template_columns": null,
            "grid_template_rows": null,
            "height": null,
            "justify_content": null,
            "justify_items": null,
            "left": null,
            "margin": null,
            "max_height": null,
            "max_width": null,
            "min_height": null,
            "min_width": null,
            "object_fit": null,
            "object_position": null,
            "order": null,
            "overflow": null,
            "overflow_x": null,
            "overflow_y": null,
            "padding": null,
            "right": null,
            "top": null,
            "visibility": null,
            "width": null
          }
        },
        "94ce74af3988415ca43a5d23accce7ca": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "DescriptionStyleModel",
          "model_module_version": "1.5.0",
          "state": {
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "DescriptionStyleModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "StyleView",
            "description_width": ""
          }
        },
        "c64aac8fce884e45a012006e2dc20402": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "HTMLModel",
          "model_module_version": "1.5.0",
          "state": {
            "_dom_classes": [],
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "HTMLModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/controls",
            "_view_module_version": "1.5.0",
            "_view_name": "HTMLView",
            "description": "",
            "description_tooltip": null,
            "layout": "IPY_MODEL_a7231aad9ce941bbbe2c5916fbb502d3",
            "placeholder": "​",
            "style": "IPY_MODEL_ccae43d897aa4185938744e0e68c8eca",
            "value": "<b><font color=\"blue\">Chatbot:</font></b>  Ashish Vaswani, Noam Shazeer, Niki Parmar, Jakob Uszkoreit, Llion Jones, Aidan N. Gomez, Łukasz Kaiser, and Illia Polosukhin created transformers."
          }
        },
        "a7231aad9ce941bbbe2c5916fbb502d3": {
          "model_module": "@jupyter-widgets/base",
          "model_name": "LayoutModel",
          "model_module_version": "1.2.0",
          "state": {
            "_model_module": "@jupyter-widgets/base",
            "_model_module_version": "1.2.0",
            "_model_name": "LayoutModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "LayoutView",
            "align_content": null,
            "align_items": null,
            "align_self": null,
            "border": null,
            "bottom": null,
            "display": null,
            "flex": null,
            "flex_flow": null,
            "grid_area": null,
            "grid_auto_columns": null,
            "grid_auto_flow": null,
            "grid_auto_rows": null,
            "grid_column": null,
            "grid_gap": null,
            "grid_row": null,
            "grid_template_areas": null,
            "grid_template_columns": null,
            "grid_template_rows": null,
            "height": null,
            "justify_content": null,
            "justify_items": null,
            "left": null,
            "margin": null,
            "max_height": null,
            "max_width": null,
            "min_height": null,
            "min_width": null,
            "object_fit": null,
            "object_position": null,
            "order": null,
            "overflow": null,
            "overflow_x": null,
            "overflow_y": null,
            "padding": null,
            "right": null,
            "top": null,
            "visibility": null,
            "width": null
          }
        },
        "ccae43d897aa4185938744e0e68c8eca": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "DescriptionStyleModel",
          "model_module_version": "1.5.0",
          "state": {
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "DescriptionStyleModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "StyleView",
            "description_width": ""
          }
        },
        "a6fc81eb00c1487f8bcf296d3f1f09ba": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "HTMLModel",
          "model_module_version": "1.5.0",
          "state": {
            "_dom_classes": [],
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "HTMLModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/controls",
            "_view_module_version": "1.5.0",
            "_view_name": "HTMLView",
            "description": "",
            "description_tooltip": null,
            "layout": "IPY_MODEL_4b2f4c5464294d38af0c4b502c3aa838",
            "placeholder": "​",
            "style": "IPY_MODEL_0548b5525848403498cfcad4fed553d3",
            "value": "<b>User:</b> Were they smart?"
          }
        },
        "4b2f4c5464294d38af0c4b502c3aa838": {
          "model_module": "@jupyter-widgets/base",
          "model_name": "LayoutModel",
          "model_module_version": "1.2.0",
          "state": {
            "_model_module": "@jupyter-widgets/base",
            "_model_module_version": "1.2.0",
            "_model_name": "LayoutModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "LayoutView",
            "align_content": null,
            "align_items": null,
            "align_self": null,
            "border": null,
            "bottom": null,
            "display": null,
            "flex": null,
            "flex_flow": null,
            "grid_area": null,
            "grid_auto_columns": null,
            "grid_auto_flow": null,
            "grid_auto_rows": null,
            "grid_column": null,
            "grid_gap": null,
            "grid_row": null,
            "grid_template_areas": null,
            "grid_template_columns": null,
            "grid_template_rows": null,
            "height": null,
            "justify_content": null,
            "justify_items": null,
            "left": null,
            "margin": null,
            "max_height": null,
            "max_width": null,
            "min_height": null,
            "min_width": null,
            "object_fit": null,
            "object_position": null,
            "order": null,
            "overflow": null,
            "overflow_x": null,
            "overflow_y": null,
            "padding": null,
            "right": null,
            "top": null,
            "visibility": null,
            "width": null
          }
        },
        "0548b5525848403498cfcad4fed553d3": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "DescriptionStyleModel",
          "model_module_version": "1.5.0",
          "state": {
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "DescriptionStyleModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "StyleView",
            "description_width": ""
          }
        },
        "5b78945e3add476b94fdbfe8eb632471": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "HTMLModel",
          "model_module_version": "1.5.0",
          "state": {
            "_dom_classes": [],
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "HTMLModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/controls",
            "_view_module_version": "1.5.0",
            "_view_name": "HTMLView",
            "description": "",
            "description_tooltip": null,
            "layout": "IPY_MODEL_ca43dc21b81546faa8cc778937b073dc",
            "placeholder": "​",
            "style": "IPY_MODEL_61c8dc3391e148e8a2467caff97392a1",
            "value": "<b><font color=\"blue\">Chatbot:</font></b>  I don't know."
          }
        },
        "ca43dc21b81546faa8cc778937b073dc": {
          "model_module": "@jupyter-widgets/base",
          "model_name": "LayoutModel",
          "model_module_version": "1.2.0",
          "state": {
            "_model_module": "@jupyter-widgets/base",
            "_model_module_version": "1.2.0",
            "_model_name": "LayoutModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "LayoutView",
            "align_content": null,
            "align_items": null,
            "align_self": null,
            "border": null,
            "bottom": null,
            "display": null,
            "flex": null,
            "flex_flow": null,
            "grid_area": null,
            "grid_auto_columns": null,
            "grid_auto_flow": null,
            "grid_auto_rows": null,
            "grid_column": null,
            "grid_gap": null,
            "grid_row": null,
            "grid_template_areas": null,
            "grid_template_columns": null,
            "grid_template_rows": null,
            "height": null,
            "justify_content": null,
            "justify_items": null,
            "left": null,
            "margin": null,
            "max_height": null,
            "max_width": null,
            "min_height": null,
            "min_width": null,
            "object_fit": null,
            "object_position": null,
            "order": null,
            "overflow": null,
            "overflow_x": null,
            "overflow_y": null,
            "padding": null,
            "right": null,
            "top": null,
            "visibility": null,
            "width": null
          }
        },
        "61c8dc3391e148e8a2467caff97392a1": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "DescriptionStyleModel",
          "model_module_version": "1.5.0",
          "state": {
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "DescriptionStyleModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "StyleView",
            "description_width": ""
          }
        },
        "9a4f5a05dbb344e897b6be9b2f7b4ad0": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "HTMLModel",
          "model_module_version": "1.5.0",
          "state": {
            "_dom_classes": [],
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "HTMLModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/controls",
            "_view_module_version": "1.5.0",
            "_view_name": "HTMLView",
            "description": "",
            "description_tooltip": null,
            "layout": "IPY_MODEL_322af3a06d1646f794f69dea2b799fcb",
            "placeholder": "​",
            "style": "IPY_MODEL_5540d70ffbeb4126ae94aed1ebaa95aa",
            "value": "<b>User:</b> I think they were"
          }
        },
        "322af3a06d1646f794f69dea2b799fcb": {
          "model_module": "@jupyter-widgets/base",
          "model_name": "LayoutModel",
          "model_module_version": "1.2.0",
          "state": {
            "_model_module": "@jupyter-widgets/base",
            "_model_module_version": "1.2.0",
            "_model_name": "LayoutModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "LayoutView",
            "align_content": null,
            "align_items": null,
            "align_self": null,
            "border": null,
            "bottom": null,
            "display": null,
            "flex": null,
            "flex_flow": null,
            "grid_area": null,
            "grid_auto_columns": null,
            "grid_auto_flow": null,
            "grid_auto_rows": null,
            "grid_column": null,
            "grid_gap": null,
            "grid_row": null,
            "grid_template_areas": null,
            "grid_template_columns": null,
            "grid_template_rows": null,
            "height": null,
            "justify_content": null,
            "justify_items": null,
            "left": null,
            "margin": null,
            "max_height": null,
            "max_width": null,
            "min_height": null,
            "min_width": null,
            "object_fit": null,
            "object_position": null,
            "order": null,
            "overflow": null,
            "overflow_x": null,
            "overflow_y": null,
            "padding": null,
            "right": null,
            "top": null,
            "visibility": null,
            "width": null
          }
        },
        "5540d70ffbeb4126ae94aed1ebaa95aa": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "DescriptionStyleModel",
          "model_module_version": "1.5.0",
          "state": {
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "DescriptionStyleModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "StyleView",
            "description_width": ""
          }
        },
        "22308835851f4231a9f4787a718d2e5c": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "HTMLModel",
          "model_module_version": "1.5.0",
          "state": {
            "_dom_classes": [],
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "HTMLModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/controls",
            "_view_module_version": "1.5.0",
            "_view_name": "HTMLView",
            "description": "",
            "description_tooltip": null,
            "layout": "IPY_MODEL_14005694bb3d4bfbabd3944d8d49fa1c",
            "placeholder": "​",
            "style": "IPY_MODEL_d2125f0742944b1f95996928c87a1a80",
            "value": "<b><font color=\"blue\">Chatbot:</font></b>  Yes, I think they were smart."
          }
        },
        "14005694bb3d4bfbabd3944d8d49fa1c": {
          "model_module": "@jupyter-widgets/base",
          "model_name": "LayoutModel",
          "model_module_version": "1.2.0",
          "state": {
            "_model_module": "@jupyter-widgets/base",
            "_model_module_version": "1.2.0",
            "_model_name": "LayoutModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "LayoutView",
            "align_content": null,
            "align_items": null,
            "align_self": null,
            "border": null,
            "bottom": null,
            "display": null,
            "flex": null,
            "flex_flow": null,
            "grid_area": null,
            "grid_auto_columns": null,
            "grid_auto_flow": null,
            "grid_auto_rows": null,
            "grid_column": null,
            "grid_gap": null,
            "grid_row": null,
            "grid_template_areas": null,
            "grid_template_columns": null,
            "grid_template_rows": null,
            "height": null,
            "justify_content": null,
            "justify_items": null,
            "left": null,
            "margin": null,
            "max_height": null,
            "max_width": null,
            "min_height": null,
            "min_width": null,
            "object_fit": null,
            "object_position": null,
            "order": null,
            "overflow": null,
            "overflow_x": null,
            "overflow_y": null,
            "padding": null,
            "right": null,
            "top": null,
            "visibility": null,
            "width": null
          }
        },
        "d2125f0742944b1f95996928c87a1a80": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "DescriptionStyleModel",
          "model_module_version": "1.5.0",
          "state": {
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "DescriptionStyleModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "StyleView",
            "description_width": ""
          }
        }
      }
    }
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "view-in-github",
        "colab_type": "text"
      },
      "source": [
        "<a href=\"https://colab.research.google.com/github/yevhen-samoilov/databaseStorage/blob/main/%F0%9F%A6%9C%F0%9F%94%97_Chat_with_PDFs_Custom_Knowledge_ChatGPT_with_LangChain.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "# **Custom Knowledge ChatGPT with LangChain - Chat with PDFs**\n",
        "\n",
        "**By Liam Ottley:**  [YouTube](https://youtube.com/@LiamOttley)\n",
        "\n",
        "\n",
        "\n",
        "\n",
        "\n",
        "0.   Installs, Imports and API Keys\n",
        "1.   Loading PDFs and chunking with LangChain\n",
        "2.   Embedding text and storing embeddings\n",
        "3.   Creating retrieval function\n",
        "4.   Creating chatbot with chat memory (OPTIONAL) \n",
        "\n",
        "\n",
        "\n",
        "\n",
        "\n",
        "\n"
      ],
      "metadata": {
        "id": "_x1GI7Fo8Y7x"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "# 0. Installs, Imports and API Keys"
      ],
      "metadata": {
        "id": "Q24Y-g6h-Bg0"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "# RUN THIS CELL FIRST!\n",
        "!pip install -q langchain==0.0.150 pypdf pandas matplotlib tiktoken textract transformers openai faiss-cpu"
      ],
      "metadata": {
        "id": "gk2J2sYYjTkM"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "import os\n",
        "import pandas as pd\n",
        "import matplotlib.pyplot as plt\n",
        "from transformers import GPT2TokenizerFast\n",
        "from langchain.document_loaders import PyPDFLoader\n",
        "from langchain.text_splitter import RecursiveCharacterTextSplitter\n",
        "from langchain.embeddings import OpenAIEmbeddings\n",
        "from langchain.vectorstores import FAISS\n",
        "from langchain.chains.question_answering import load_qa_chain\n",
        "from langchain.llms import OpenAI\n",
        "from langchain.chains import ConversationalRetrievalChain"
      ],
      "metadata": {
        "id": "l-uszlwN641q"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "os.environ[\"OPENAI_API_KEY\"] = \"{YOURAPIKEY}\""
      ],
      "metadata": {
        "id": "E2Buv5Y0uFr8"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "# 1. Loading PDFs and chunking with LangChain"
      ],
      "metadata": {
        "id": "RLULMPXa-Hu8"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "# You MUST add your PDF to local files in this notebook (folder icon on left hand side of screen)\n",
        "\n",
        "# Simple method - Split by pages \n",
        "loader = PyPDFLoader(\"./attention_is_all_you_need.pdf\")\n",
        "pages = loader.load_and_split()\n",
        "print(pages[0])\n",
        "\n",
        "# SKIP TO STEP 2 IF YOU'RE USING THIS METHOD\n",
        "chunks = pages"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "KH546j3nkFwX",
        "outputId": "fc5bb7da-59ac-49d4-8ac5-2cfe73483225"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "page_content='Attention Is All You Need\\nAshish Vaswani\\x03\\nGoogle Brain\\navaswani@google.comNoam Shazeer\\x03\\nGoogle Brain\\nnoam@google.comNiki Parmar\\x03\\nGoogle Research\\nnikip@google.comJakob Uszkoreit\\x03\\nGoogle Research\\nusz@google.com\\nLlion Jones\\x03\\nGoogle Research\\nllion@google.comAidan N. Gomez\\x03y\\nUniversity of Toronto\\naidan@cs.toronto.eduŁukasz Kaiser\\x03\\nGoogle Brain\\nlukaszkaiser@google.com\\nIllia Polosukhin\\x03z\\nillia.polosukhin@gmail.com\\nAbstract\\nThe dominant sequence transduction models are based on complex recurrent or\\nconvolutional neural networks that include an encoder and a decoder. The best\\nperforming models also connect the encoder and decoder through an attention\\nmechanism. We propose a new simple network architecture, the Transformer,\\nbased solely on attention mechanisms, dispensing with recurrence and convolutions\\nentirely. Experiments on two machine translation tasks show these models to\\nbe superior in quality while being more parallelizable and requiring signiﬁcantly\\nless time to train. Our model achieves 28.4 BLEU on the WMT 2014 English-\\nto-German translation task, improving over the existing best results, including\\nensembles, by over 2 BLEU. On the WMT 2014 English-to-French translation task,\\nour model establishes a new single-model state-of-the-art BLEU score of 41.8 after\\ntraining for 3.5 days on eight GPUs, a small fraction of the training costs of the\\nbest models from the literature. We show that the Transformer generalizes well to\\nother tasks by applying it successfully to English constituency parsing both with\\nlarge and limited training data.\\n1 Introduction\\nRecurrent neural networks, long short-term memory [ 13] and gated recurrent [ 7] neural networks\\nin particular, have been ﬁrmly established as state of the art approaches in sequence modeling and\\n\\x03Equal contribution. Listing order is random. Jakob proposed replacing RNNs with self-attention and started\\nthe effort to evaluate this idea. Ashish, with Illia, designed and implemented the ﬁrst Transformer models and\\nhas been crucially involved in every aspect of this work. Noam proposed scaled dot-product attention, multi-head\\nattention and the parameter-free position representation and became the other person involved in nearly every\\ndetail. Niki designed, implemented, tuned and evaluated countless model variants in our original codebase and\\ntensor2tensor. Llion also experimented with novel model variants, was responsible for our initial codebase, and\\nefﬁcient inference and visualizations. Lukasz and Aidan spent countless long days designing various parts of and\\nimplementing tensor2tensor, replacing our earlier codebase, greatly improving results and massively accelerating\\nour research.\\nyWork performed while at Google Brain.\\nzWork performed while at Google Research.\\n31st Conference on Neural Information Processing Systems (NIPS 2017), Long Beach, CA, USA.arXiv:1706.03762v5  [cs.CL]  6 Dec 2017' metadata={'source': './attention_is_all_you_need.pdf', 'page': 0}\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# Advanced method - Split by chunk\n",
        "\n",
        "# Step 1: Convert PDF to text\n",
        "import textract\n",
        "doc = textract.process(\"./attention_is_all_you_need.pdf\")\n",
        "\n",
        "# Step 2: Save to .txt and reopen (helps prevent issues)\n",
        "with open('attention_is_all_you_need.txt', 'w') as f:\n",
        "    f.write(doc.decode('utf-8'))\n",
        "\n",
        "with open('attention_is_all_you_need.txt', 'r') as f:\n",
        "    text = f.read()\n",
        "\n",
        "# Step 3: Create function to count tokens\n",
        "tokenizer = GPT2TokenizerFast.from_pretrained(\"gpt2\")\n",
        "\n",
        "def count_tokens(text: str) -> int:\n",
        "    return len(tokenizer.encode(text))\n",
        "\n",
        "# Step 4: Split text into chunks\n",
        "text_splitter = RecursiveCharacterTextSplitter(\n",
        "    # Set a really small chunk size, just to show.\n",
        "    chunk_size = 512,\n",
        "    chunk_overlap  = 24,\n",
        "    length_function = count_tokens,\n",
        ")\n",
        "\n",
        "chunks = text_splitter.create_documents([text])"
      ],
      "metadata": {
        "id": "iADY2CXNlNq9"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "# Result is many LangChain 'Documents' around 500 tokens or less (Recursive splitter sometimes allows more tokens to retain context)\n",
        "type(chunks[0]) "
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "KQ_gDkwep4q7",
        "outputId": "ef5323ce-2ea2-4a44-f8cd-31f93d066112"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "langchain.schema.Document"
            ]
          },
          "metadata": {},
          "execution_count": 93
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# Quick data visualization to ensure chunking was successful\n",
        "\n",
        "# Create a list of token counts\n",
        "token_counts = [count_tokens(chunk.page_content) for chunk in chunks]\n",
        "\n",
        "# Create a DataFrame from the token counts\n",
        "df = pd.DataFrame({'Token Count': token_counts})\n",
        "\n",
        "# Create a histogram of the token count distribution\n",
        "df.hist(bins=40, )\n",
        "\n",
        "# Show the plot\n",
        "plt.show()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 452
        },
        "id": "fK31bxDOpz1l",
        "outputId": "ad765c54-d21b-43ad-ee6a-1efe0c3963ca"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 640x480 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "# 2. Embed text and store embeddings"
      ],
      "metadata": {
        "id": "_IlznUDK-i2m"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "# Get embedding model\n",
        "embeddings = OpenAIEmbeddings()\n",
        "\n",
        "# Create vector database\n",
        "db = FAISS.from_documents(chunks, embeddings)"
      ],
      "metadata": {
        "id": "92ObhTAKnZzQ"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "# 3. Setup retrieval function"
      ],
      "metadata": {
        "id": "2LPwdGDP-nPO"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "# Check similarity search is working\n",
        "query = \"Who created transformers?\"\n",
        "docs = db.similarity_search(query)\n",
        "docs[0]"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "RWP92zGg5Nb_",
        "outputId": "6c698c0c-d895-4aac-ff7c-15c557435a9d"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "Document(page_content='7 Conclusion\\n\\nIn this work, we presented the Transformer, the ﬁrst sequence transduction model based entirely on\\nattention, replacing the recurrent layers most commonly used in encoder-decoder architectures with\\nmulti-headed self-attention.\\nFor translation tasks, the Transformer can be trained signiﬁcantly faster than architectures based\\non recurrent or convolutional layers. On both WMT 2014 English-to-German and WMT 2014\\nEnglish-to-French translation tasks, we achieve a new state of the art. In the former task our best\\nmodel outperforms even all previously reported ensembles.\\nWe are excited about the future of attention-based models and plan to apply them to other tasks. We\\nplan to extend the Transformer to problems involving input and output modalities other than text and\\nto investigate local, restricted attention mechanisms to efﬁciently handle large inputs and outputs\\nsuch as images, audio and video. Making generation less sequential is another research goals of ours.\\nThe code we used to train and evaluate our models is available at https://github.com/\\ntensorflow/tensor2tensor.\\n\\nAcknowledgements We are grateful to Nal Kalchbrenner and Stephan Gouws for their fruitful\\ncomments, corrections and inspiration.\\n\\nReferences\\n[1] Jimmy Lei Ba, Jamie Ryan Kiros, and Geoffrey E Hinton. Layer normalization. arXiv preprint\\n\\narXiv:1607.06450, 2016.\\n\\n[2] Dzmitry Bahdanau, Kyunghyun Cho, and Yoshua Bengio. Neural machine translation by jointly\\n\\nlearning to align and translate. CoRR, abs/1409.0473, 2014.\\n\\n[3] Denny Britz, Anna Goldie, Minh-Thang Luong, and Quoc V. Le. Massive exploration of neural\\n\\nmachine translation architectures. CoRR, abs/1703.03906, 2017.\\n\\n[4] Jianpeng Cheng, Li Dong, and Mirella Lapata. Long short-term memory-networks for machine\\n\\nreading. arXiv preprint arXiv:1601.06733, 2016.', metadata={})"
            ]
          },
          "metadata": {},
          "execution_count": 98
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# Create QA chain to integrate similarity search with user queries (answer query from knowledge base)\n",
        "\n",
        "chain = load_qa_chain(OpenAI(temperature=0), chain_type=\"stuff\")\n",
        "\n",
        "query = \"Who created transformers?\"\n",
        "docs = db.similarity_search(query)\n",
        "\n",
        "chain.run(input_documents=docs, question=query)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 53
        },
        "id": "1Kv_sM8G5qAo",
        "outputId": "00171d21-312a-455e-b361-2306c6dc4b75"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "' Ashish Vaswani, Noam Shazeer, Niki Parmar, Jakob Uszkoreit, Llion Jones, Aidan N. Gomez, Łukasz Kaiser, and Illia Polosukhin created transformers.'"
            ],
            "application/vnd.google.colaboratory.intrinsic+json": {
              "type": "string"
            }
          },
          "metadata": {},
          "execution_count": 99
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "# 5. Create chatbot with chat memory (OPTIONAL) "
      ],
      "metadata": {
        "id": "U_nH1qoL-w--"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "from IPython.display import display\n",
        "import ipywidgets as widgets\n",
        "\n",
        "# Create conversation chain that uses our vectordb as retriver, this also allows for chat history management\n",
        "qa = ConversationalRetrievalChain.from_llm(OpenAI(temperature=0.1), db.as_retriever())"
      ],
      "metadata": {
        "id": "evF7_Dyhtcaf"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "chat_history = []\n",
        "\n",
        "def on_submit(_):\n",
        "    query = input_box.value\n",
        "    input_box.value = \"\"\n",
        "    \n",
        "    if query.lower() == 'exit':\n",
        "        print(\"Thank you for using the State of the Union chatbot!\")\n",
        "        return\n",
        "    \n",
        "    result = qa({\"question\": query, \"chat_history\": chat_history})\n",
        "    chat_history.append((query, result['answer']))\n",
        "    \n",
        "    display(widgets.HTML(f'<b>User:</b> {query}'))\n",
        "    display(widgets.HTML(f'<b><font color=\"blue\">Chatbot:</font></b> {result[\"answer\"]}'))\n",
        "\n",
        "print(\"Welcome to the Transformers chatbot! Type 'exit' to stop.\")\n",
        "\n",
        "input_box = widgets.Text(placeholder='Please enter your question:')\n",
        "input_box.on_submit(on_submit)\n",
        "\n",
        "display(input_box)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 275,
          "referenced_widgets": [
            "bf4eac20add04b57a1af538f80f2980c",
            "867b911c7aac4dc7bdf6923d3a464c1e",
            "a32ffe2bd6974b6eaa8a3081a8c1816b",
            "7a260cc663614805b499690a5a2cb5be",
            "93571528b372443592d502bebd336ca0",
            "94ce74af3988415ca43a5d23accce7ca",
            "c64aac8fce884e45a012006e2dc20402",
            "a7231aad9ce941bbbe2c5916fbb502d3",
            "ccae43d897aa4185938744e0e68c8eca",
            "a6fc81eb00c1487f8bcf296d3f1f09ba",
            "4b2f4c5464294d38af0c4b502c3aa838",
            "0548b5525848403498cfcad4fed553d3",
            "5b78945e3add476b94fdbfe8eb632471",
            "ca43dc21b81546faa8cc778937b073dc",
            "61c8dc3391e148e8a2467caff97392a1",
            "9a4f5a05dbb344e897b6be9b2f7b4ad0",
            "322af3a06d1646f794f69dea2b799fcb",
            "5540d70ffbeb4126ae94aed1ebaa95aa",
            "22308835851f4231a9f4787a718d2e5c",
            "14005694bb3d4bfbabd3944d8d49fa1c",
            "d2125f0742944b1f95996928c87a1a80"
          ]
        },
        "id": "-pHw5siewPNt",
        "outputId": "e3fb32a5-84fc-47a1-84cd-df0c763f7720"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Welcome to the Transformers chatbot! Type 'exit' to stop.\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "Text(value='', placeholder='Please enter your question:')"
            ],
            "application/vnd.jupyter.widget-view+json": {
              "version_major": 2,
              "version_minor": 0,
              "model_id": "bf4eac20add04b57a1af538f80f2980c"
            }
          },
          "metadata": {}
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "HTML(value='<b>User:</b> Who created transformers?')"
            ],
            "application/vnd.jupyter.widget-view+json": {
              "version_major": 2,
              "version_minor": 0,
              "model_id": "7a260cc663614805b499690a5a2cb5be"
            }
          },
          "metadata": {}
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "HTML(value='<b><font color=\"blue\">Chatbot:</font></b>  Ashish Vaswani, Noam Shazeer, Niki Parmar, Jakob Uszkor…"
            ],
            "application/vnd.jupyter.widget-view+json": {
              "version_major": 2,
              "version_minor": 0,
              "model_id": "c64aac8fce884e45a012006e2dc20402"
            }
          },
          "metadata": {}
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "HTML(value='<b>User:</b> Were they smart?')"
            ],
            "application/vnd.jupyter.widget-view+json": {
              "version_major": 2,
              "version_minor": 0,
              "model_id": "a6fc81eb00c1487f8bcf296d3f1f09ba"
            }
          },
          "metadata": {}
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "HTML(value='<b><font color=\"blue\">Chatbot:</font></b>  I don\\'t know.')"
            ],
            "application/vnd.jupyter.widget-view+json": {
              "version_major": 2,
              "version_minor": 0,
              "model_id": "5b78945e3add476b94fdbfe8eb632471"
            }
          },
          "metadata": {}
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "HTML(value='<b>User:</b> I think they were')"
            ],
            "application/vnd.jupyter.widget-view+json": {
              "version_major": 2,
              "version_minor": 0,
              "model_id": "9a4f5a05dbb344e897b6be9b2f7b4ad0"
            }
          },
          "metadata": {}
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "HTML(value='<b><font color=\"blue\">Chatbot:</font></b>  Yes, I think they were smart.')"
            ],
            "application/vnd.jupyter.widget-view+json": {
              "version_major": 2,
              "version_minor": 0,
              "model_id": "22308835851f4231a9f4787a718d2e5c"
            }
          },
          "metadata": {}
        }
      ]
    }
  ]
}