reponse-api-rag-file-search.ipynb

reponse-api-rag-file-search.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "provenance": [],
      "include_colab_link": true
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    },
    "language_info": {
      "name": "python"
    }
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "view-in-github",
        "colab_type": "text"
      },
      "source": [
        "<a href=\"https://colab.research.google.com/gist/FinanceData/d7d2560824afa8c3f34c31e9ee7cd040/reponse-api-rag-file-search.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "# Reponse API - RAG 구현\n",
        "\n",
        "\n",
        "**2024~2025 [FinanceData.KR]()**"
      ],
      "metadata": {
        "id": "LOooAvo9JCPv"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "import os\n",
        "os.environ[\"OPENAI_API_KEY\"] = \"***********************************************\"\n"
      ],
      "metadata": {
        "id": "7eLEt9wE66sB"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "## 시총상위 10개사 사업보고서(2024년 반기)"
      ],
      "metadata": {
        "id": "rnF4L61eLyxP"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "!gdown \"1T5tcPTQSi_jjgiP7HMwzDdvJul1zE4Vt\""
      ],
      "metadata": {
        "id": "04kEt_grLsso"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "!unzip \"2024_반기사업보고서.zip\""
      ],
      "metadata": {
        "id": "EcTa8TlhLsnc"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "import glob\n",
        "\n",
        "pdf_files = glob.glob(\"*.pdf\")\n",
        "pdf_files"
      ],
      "metadata": {
        "collapsed": true,
        "id": "Qwbwl3wBL69y"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "oZKfngV5lJZJ",
        "collapsed": true
      },
      "outputs": [],
      "source": [
        "!pip install -q openai"
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "from openai import OpenAI\n",
        "client = OpenAI()"
      ],
      "metadata": {
        "id": "r7OHRdioOaDt"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "## 1단계: 벡터스토어 생성및 파일 업로드\n"
      ],
      "metadata": {
        "id": "Hinhl7q1SL6F"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "from openai import OpenAI\n",
        "client = OpenAI()\n",
        "\n",
        "vector_store = client.vector_stores.create(        # Create vector store\n",
        "    name=\"dart-report\",\n",
        ")"
      ],
      "metadata": {
        "id": "J-xAKtB77ZMO"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "vector_store"
      ],
      "metadata": {
        "id": "BXKHpN5s8lZr"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "pdf_files"
      ],
      "metadata": {
        "id": "jVnBz9WP-Cbb"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "for pdf_file in pdf_files:\n",
        "    client.vector_stores.files.upload_and_poll(\n",
        "        vector_store_id=vector_store.id,\n",
        "        file=open(pdf_file, \"rb\")\n",
        "    )"
      ],
      "metadata": {
        "id": "evxhXXF4-Bnm"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "user_query = \"셀트리온의 제무제표를 표로 정리해서 출력해줘\"\n",
        "\n",
        "results = client.vector_stores.search(\n",
        "    vector_store_id=vector_store.id,\n",
        "    query=user_query,\n",
        ")\n",
        "\n",
        "results.to_dict() # 확인"
      ],
      "metadata": {
        "id": "hxltltLx-R96"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "## 2단계: 벡터스토어에 RAG 수행"
      ],
      "metadata": {
        "id": "iPcrDre2-aCt"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "from openai import OpenAI\n",
        "client = OpenAI()\n",
        "\n",
        "response = client.responses.create(\n",
        "    model=\"gpt-4o-mini\",\n",
        "    input=\"최근 3년 셀트리온의 제무제표를 상세하게 출력해줘\",\n",
        "    tools=[{\n",
        "        \"type\": \"file_search\",\n",
        "        \"vector_store_ids\": [vector_store.id]\n",
        "    }]\n",
        ")\n",
        "\n",
        "print(response.output_text)"
      ],
      "metadata": {
        "id": "-y8M7D0i-Z9b"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "from openai import OpenAI\n",
        "client = OpenAI()\n",
        "\n",
        "response = client.responses.create(\n",
        "    model=\"gpt-4o-mini\",\n",
        "    input=\"현대자동차의 재무제표를 투자 관점에서 평가해줘, 출처도 함께 표기해줘\",\n",
        "    tools=[{\n",
        "        \"type\": \"file_search\",\n",
        "        \"vector_store_ids\": [vector_store.id]\n",
        "    }]\n",
        ")\n",
        "\n",
        "print(response.output_text)"
      ],
      "metadata": {
        "id": "FP8lQSqGAavB"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "\n",
        "**2024~2025 [FinanceData.KR]()**"
      ],
      "metadata": {
        "id": "YzrtfGI0XdNG"
      }
    }
  ]
}