LLM_Based_Egogram_test.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "provenance": [],
      "authorship_tag": "ABX9TyM1AiZ9xNSRphmuVbxqcjuN",
      "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/github/hosseinhimself/LLM-Based-Egogram-Test/blob/main/LLM_Based_Egogram_test.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 1,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "cellView": "form",
        "id": "HfEmO4AnQ838",
        "outputId": "c27ea0a0-47da-435e-b87b-ae22423bda4d"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "\u001b[2K     \u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m810.5/810.5 kB\u001b[0m \u001b[31m3.3 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0m\n",
            "\u001b[2K     \u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m1.8/1.8 MB\u001b[0m \u001b[31m3.7 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0m\n",
            "\u001b[2K     \u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m269.1/269.1 kB\u001b[0m \u001b[31m2.9 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0m\n",
            "\u001b[2K     \u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m71.6/71.6 kB\u001b[0m \u001b[31m2.6 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0m\n",
            "\u001b[2K     \u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m49.4/49.4 kB\u001b[0m \u001b[31m2.1 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0m\n",
            "\u001b[2K     \u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m53.0/53.0 kB\u001b[0m \u001b[31m1.9 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0m\n",
            "\u001b[2K     \u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m138.5/138.5 kB\u001b[0m \u001b[31m773.9 kB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0m\n",
            "\u001b[2K     \u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m137.4/137.4 kB\u001b[0m \u001b[31m2.1 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0m\n",
            "\u001b[?25h"
          ]
        }
      ],
      "source": [
        "# @title Requirements\n",
        "! pip install -q -U langchain\n",
        "! pip install -q langchain-google-genai\n",
        "\n",
        "from langchain import PromptTemplate\n",
        "from langchain.document_loaders import WebBaseLoader\n",
        "from langchain.schema import StrOutputParser\n",
        "from langchain.schema.prompt_template import format_document\n",
        "from langchain.chains import LLMChain\n",
        "from langchain_google_genai import ChatGoogleGenerativeAI\n",
        "\n",
        "import ast\n",
        "import os\n",
        "import getpass\n",
        "import matplotlib.pyplot as plt\n"
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# @title Set API Key\n",
        "api_key = getpass.getpass('Gemini API Key:')\n",
        "os.environ['GOOGLE_API_KEY'] = api_key"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "cellView": "form",
        "id": "MTQr3USVRJHp",
        "outputId": "b9389364-f232-4ef3-d2ed-50c9edcc40f7"
      },
      "execution_count": 36,
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Gemini API Key:··········\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# @title Prompts\n",
        "initial_task_prompt = '''\n",
        "  You are a psychologist and you have the duty to ask questions in order to find egogram (child, adult, parent) of the person answering the questions can be identified.\n",
        "  Ask {number_of_questions} questions, where each question has an answer between numbers 1 to 5, where 1 is very low, 2 is low, 3 is average, 4 is high, and 5 is very high.\n",
        "  the questions must be related to emotions, personality traits and also about emotions in hypothetical situations.\n",
        "  give me the questions in this format:\n",
        "  Q1:<<question text>>\\nQ2:<<question text>>\\n...\\nQ20:<<question text>>\n",
        "  just write questions and dont add any other informations\n",
        "  '''\n",
        "\n",
        "welcome = '''\n",
        "  **Welcome!**\n",
        "  To determine your egogram, please answer the following questions by selecting a number between 1 and 5, where:\n",
        "  1 = Very low\n",
        "  2 = Low\n",
        "  3 = Average\n",
        "  4 = High\n",
        "  5 = Very high\n",
        "  '''\n",
        "\n",
        "results_prompt = '''\n",
        "  You are a psychologist and your task is to determine the percentage of the user's egogram (child, adult, parent) based on the questions that have already been asked. \\\n",
        "  The questions are related to emotions, personality traits and also about emotions in hypothetical situations. their answers are from 1 to 5 as follows:\\\n",
        "  1 = Very low \\\n",
        "  2 = Low \\\n",
        "  3 = Average \\\n",
        "  4 = High \\\n",
        "  5 = Very high \\\n",
        "  Please disregard any questions that have not been answered or have received answers unrelated to the topic.\n",
        "  The questions and their answers are given in the following format as a python dictionary:\n",
        "  '<<question>>':'<<answer>>',...\n",
        "  Return the answer to me exactly in the following format. ego names are string and in '' mark and numbers must be float without \"%\". and do not write anything else extra:\n",
        "  'Child-Ego':<<result in percent>>, 'Parent-Ego':<<result in percent>>, 'Adult-Ego':<<result in percent>>\n",
        "\n",
        "\n",
        "  questions and answers:\n",
        "\n",
        "  {answers}\n",
        "\n",
        "  '''\n"
      ],
      "metadata": {
        "cellView": "form",
        "id": "cZVhstjdRRSP"
      },
      "execution_count": 28,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "# @title Run the test\n",
        "llm_model = ChatGoogleGenerativeAI(model=\"gemini-pro\",\n",
        "                 temperature=0.7, top_p=0.85)\n",
        "\n",
        "def generate_questions(number_of_questions):\n",
        "  prompt = PromptTemplate(input_variables=[\"number_of_questions\"], template = initial_task_prompt)\n",
        "  # define chain\n",
        "  chain = LLMChain(\n",
        "      llm=llm_model,\n",
        "      prompt=prompt,\n",
        "  )\n",
        "  questions = chain.run({'number_of_questions':number_of_questions})\n",
        "  return questions.split('\\n')\n",
        "\n",
        "def show_results(answers):\n",
        "  res_prompt = PromptTemplate(input_variables=['answers'], template = results_prompt)\n",
        "  res_chain = LLMChain(\n",
        "    llm=llm_model,\n",
        "    prompt=res_prompt,\n",
        "  )\n",
        "  return res_chain.run({'answers':answers})\n",
        "\n",
        "def main_test():\n",
        "  print(welcome)\n",
        "  questions = generate_questions(10)\n",
        "  answers = {}\n",
        "  for q in questions:\n",
        "    answers[q] = input(f'{q} ')\n",
        "\n",
        "  print()\n",
        "  result = show_results(answers)\n",
        "\n",
        "  answers_dict = ast.literal_eval(f\"{{{result}}}\")\n",
        "\n",
        "  # Extract keys (labels) and values (sizes)\n",
        "  labels = list(answers_dict.keys())\n",
        "  sizes = list(answers_dict.values())\n",
        "\n",
        "  # Plotting the pie chart\n",
        "  plt.figure(figsize=(4, 4))  # Define the figure size (optional)\n",
        "  plt.pie(sizes, labels=labels, autopct='%1.1f%%', startangle=140)\n",
        "\n",
        "  # Add a title\n",
        "  plt.title('Your Egogram Chart')\n",
        "\n",
        "  # Show the plot\n",
        "  plt.show()\n",
        "\n",
        "main_test()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 714
        },
        "cellView": "form",
        "id": "qJ1b6wiIRVhO",
        "outputId": "e3fee360-e7f6-445f-9f9a-9c8ed7eec7a4"
      },
      "execution_count": 37,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "\n",
            "  **Welcome!**\n",
            "  To determine your egogram, please answer the following questions by selecting a number between 1 and 5, where:\n",
            "  1 = Very low\n",
            "  2 = Low\n",
            "  3 = Average\n",
            "  4 = High\n",
            "  5 = Very high\n",
            "  \n",
            "Q1: How often do you feel overwhelmed by your emotions? 5\n",
            "Q2: Do you consider yourself to be a confident person? 4\n",
            "Q3: How do you react when someone criticizes you? 3\n",
            "Q4: Are you comfortable expressing your feelings to others? 2\n",
            "Q5: How often do you experience feelings of guilt or shame? 3\n",
            "Q6: How do you cope with stress? 4\n",
            "Q7: If you were in a situation where you felt threatened, how likely would you be to stand up for yourself? 2\n",
            "Q8: How often do you feel like you need to please others? 1\n",
            "Q9: How well do you control your impulses? 4\n",
            "Q10: If you made a mistake, how likely would you be to apologize? 3\n",
            "\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 400x400 with 1 Axes>"
            ],
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAZgAAAFeCAYAAAC4ign8AAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjcuMSwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy/bCgiHAAAACXBIWXMAAA9hAAAPYQGoP6dpAABL80lEQVR4nO3dd3gU5doG8Hu2t/TeSEgCofceOkE6gqKiIoRioQRQsOsBEfFY8KAiqEcOiIUP1ANyFCkhdOkQSghIICGUFNLrbrbM90dkYUkg2ZDZd8vzu669ILOzM/cGNndm3ikcz/M8CCGEkEYmYh2AEEKIc6KCIYQQIggqGEIIIYKggiGEECIIKhhCCCGCoIIhhBAiCCoYQgghgqCCIYQQIggqGEIIIYKggiGEAADWrFkDjuNw7Ngx1lGIk6CCcWHDhg2Dl5cXcnJyajxXXFyMoKAgdO/eHSaTiUG6av379wfHcbU+WrRowSyXo9m4cSOGDRsGX19fyGQyBAcH4/HHH0dSUhKzTCtWrMCaNWuYrZ8IT8I6AGFnxYoVaNOmDV588UX8+OOPFs+98cYbyMvLw9atWyESsf09JDQ0FO+//36N6R4eHgzSOBae5zFlyhSsWbMGHTt2xEsvvYTAwEBkZWVh48aNGDRoEA4cOIBevXrZPNuKFSvg6+uL+Ph4m6+b2AYVjAtr2rQpFixYgFdffRXx8fF46KGHAABHjx7Fl19+ifnz56N9+/aCZjCZTKiqqoJCobjnPB4eHpgwYYKgORpbeXk51Go16xhYunQp1qxZg7lz5+KTTz4Bx3Hm595880189913kEhs+2OgoqICKpXKpuskjPDEpen1er5du3Z8VFQUX1lZyRsMBr5Tp05806ZN+fLycn7nzp187969eZVKxXt4ePCjR4/mz507Z7GMSZMm8eHh4TWWvWDBAv7u/2IA+JkzZ/Lff/8936pVK14ikfAbN268Z75+/frxrVu3rtd72bVrF9+5c2deLpfzkZGR/JdffllrBr1ezy9atIiPjIzkZTIZHx4ezr/++uu8Vqu1mM9oNPILFizgg4KCeKVSyffv359PSUnhw8PD+UmTJpnnW716NQ+A3717Nz99+nTez8+P9/T05Hme5zMyMvjp06fzzZs35xUKBe/t7c2PGzeOT09Pt1jXrWXs27ePT0hI4H19fXkPDw/+ueee43U6HV9YWMg/88wzvKenJ+/p6cm//PLLvMlkuu/3o6Kigvf29uZbtGjBGwyGOr9/tzLs37+ff/HFF3lfX19epVLxY8aM4XNzcy3m3bRpEz98+HA+KCiIl8lkfGRkJL9o0aIa67n173fs2DG+T58+vFKp5OfMmcOHh4fzACwe/fr1qzMjcSy0BePiJBIJvv76a/Tq1Qvvvvsu/P39ceLECWzduhV//vknhg0bhsjISCxcuBCVlZX4/PPPERsbixMnTiAiIqJB60xKSsKGDRswa9Ys+Pr61rkco9GIvLy8GtOVSqV5K+HkyZMYOnQogoKC8M4778BoNGLRokXw8/Or8bpp06bh22+/xbhx4zBv3jwcPnwY77//PlJTU7Fx40bzfK+//jo+/PBDjBo1CkOGDMGpU6cwZMgQaLXaWnPOmDEDfn5++Mc//oHy8nIA1VuDf/75J8aPH4/Q0FBkZGRg5cqV6N+/P86dO1fjN/mEhAQEBgbinXfewaFDh/D111/D09MTf/75J5o0aYIlS5Zgy5Yt+Oijj9CmTRtMnDjxnt+3/fv3o6CgAHPnzoVYLL7v9/juDF5eXliwYAEyMjKwbNkyzJo1C+vXrzfPs2bNGmg0Grz00kvQaDRISkrCP/7xD5SUlOCjjz6yWF5+fj6GDRuG8ePHY8KECQgICED//v2RkJAAjUaDN998EwAQEBBQ74zEQbBuOGIfZs2axUulUl6j0fBPPvkkz/M836FDB97f35/Pz883z3fq1CleJBLxEydONE+zdgtGJBLxKSkp9crVr1+/Gr/p3no8//zz5vlGjRrFq1Qq/vr16+ZpFy9e5CUSiUWG5ORkHgA/bdo0i/XMnz+fB8AnJSXxPM/z2dnZvEQi4ceMGWMx38KFC3kAtW7B9O7du8Zv8BUVFTXe08GDB3kA/Nq1a2ssY8iQIRZbJj179uQ5juNfeOEF8zSDwcCHhobW+Rv/p59+ygO47xbinW5liIuLs8jw4osv8mKxmC8qKrrv+3r++ed5lUplsSV469/vyy+/rDF/69ataavFydFRZAQA8N5778HHxwcikQj/+te/kJWVheTkZMTHx8Pb29s8X7t27TB48GBs2bKlwevq168fWrVqVe/5IyIisGPHjhqPuXPnAqjewklMTMSYMWMQHBxsfl10dDSGDRtmsaxbuV966SWL6fPmzQMA/P777wCAnTt3wmAwYMaMGRbzJSQk3DPns88+W2NLQalUmv+u1+uRn5+P6OhoeHp64sSJEzWWMXXqVItxku7du4PneUydOtU8TSwWo0uXLrh8+fI9swBASUkJAMDNze2+893tueees8jQp08fGI1GXLlypdb3VVpairy8PPTp0wcVFRU4f/68xfLkcjkmT55sVQbiHGgXGQEAuLu7IyYmBnl5eQgICMChQ4cAADExMTXmbdmyJbZt29bggeymTZtaNb9arUZcXNw9n8/NzUVlZSWio6NrPHf3tCtXrkAkEtWYHhgYCE9PT/MP0Vt/3j2ft7c3vLy8as1R2/uqrKzE+++/j9WrV+P69evg77iBbHFxcY35mzRpYvH1rSPlwsLCakwvLCysNcct7u7uAKoLwBp3Z7j1fu9cX0pKCt566y0kJSWZi+yWu99XSEgIZDKZVRmIc6CCIQ/szt9272Q0Gmudfudvv6zcK/ODqO19JSQkYPXq1Zg7dy569uwJDw8PcByH8ePH13p+0b3GSmqbztdxt/Nb5wmdOXMGY8aMqcc7uH+GW+srKipCv3794O7ujkWLFiEqKgoKhQInTpzAq6++WuN92cO/N2GDCobUKjw8HABw4cKFGs+dP38evr6+5q0XLy8vFBUV1Zjvzl0qQvL394dCoUBaWlqN5+6eFh4eDpPJhIsXL6Jly5bm6Tk5OSgqKjK/71t/pqWlWWyZ5Ofn17nlcKeff/4ZkyZNwtKlS83TtFptrd+vxta7d294eXlh3bp1eOONN6wa6L+f3bt3Iz8/H//973/Rt29f8/T09HSrliNEyRP7QmMwpFZBQUHo0KEDvv32W4sfhmfPnsX27dsxfPhw87SoqCgUFxfj9OnT5mm3TuSzBbFYjLi4OGzatAk3btwwT09LS8Mff/xhMe+t3MuWLbOY/sknnwAARowYAQAYNGgQJBIJVq5caTHf8uXLrc5295bG559/fs+tu8akUqnw6quvIjU1Fa+++mqtWzzff/89jhw5YtVybxXVncurqqrCihUrrFqOWq22SdESdmgLhtzTRx99hGHDhqFnz56YOnWq+TBlDw8PLFy40Dzf+PHj8eqrr2Ls2LGYPXs2KioqsHLlSjRv3rzWgWxrFRcX4/vvv6/1uVsnYC5cuBDbt29HbGwspk+fDqPRiOXLl6NNmzZITk42z9++fXtMmjQJX3/9tXlXz5EjR/Dtt99izJgxGDBgAIDqQ2bnzJmDpUuXYvTo0Rg6dChOnTqFP/74A76+vvX+7XvkyJH47rvv4OHhgVatWuHgwYNITEyEj4/Pg31T6unll19GSkoKli5dil27dmHcuHEIDAxEdnY2Nm3ahCNHjuDPP/+0apm9evWCl5cXJk2ahNmzZ4PjOHz33Xd17rK7W+fOnbFy5UosXrwY0dHR8Pf3x8CBA61aBrFz7A5gI/amtpMaExMT+djYWF6pVPLu7u78qFGjapxoyfM8v337dr5Nmza8TCbjY2Ji+O+///6+J1pakwn3OEz57mXv3LmT79ixIy+TyfioqCj+m2++4efNm8crFAqL+fR6Pf/OO+/wTZs25aVSKR8WFlbriZYGg4F/++23+cDAQF6pVPIDBw7kU1NTeR8fH4vDhm8d3nv06NEa+QsLC/nJkyfzvr6+vEaj4YcMGcKfP3/+nidr3r2MW9/DmzdvWkyfNGkSr1ar6/19/Pnnn/mHHnqI9/b25iUSCR8UFMQ/8cQT/O7du+vMsGvXLh4Av2vXLvO0AwcO8D169OCVSiUfHBzMv/LKK/y2bdtqzHe/E2Wzs7P5ESNG8G5ubnSipZPieN7KXzsIcSBjxoxBSkoKLl682CjLKyoqgpeXFxYvXmw+QZAQUjsagyFOo7Ky0uLrixcvYsuWLejfv3+jLA+4PXbT0GUS4kpoC4Y4jaCgIMTHxyMyMhJXrlzBypUrodPpcPLkSTRr1szq5a1ZswZr1qzB8OHDodFosH//fqxbtw4PPfQQtm3bJsA7IMS50CA/cRpDhw7FunXrkJ2dDblcjp49e2LJkiUNKheg+qoFEokEH374IUpKSswD/4sXL27k5IQ4J9qCIYQQIggagyGEECIIKhhCCCGCoIIhhBAiCCoYQgghgqCCIYQQIggqGEIIIYKggiGEECIIKhhCCCGCoIIhhBAiCCoYQgghgqCCIYQQIggqGEIIIYKggiGEECIIKhhCCCGCoIIhhBAiCCoYQgghgqCCIYQQIggqGEIIIYKggiGEECIIKhhCCCGCoIIhhBAiCCoYQgghgqCCIYQQIggqGEIIIYKggiGEECIIKhhCCCGCoIIhhBAiCCoYQgghgqCCIYQQIggqGEIIIYKggiGEECIIKhhCCCGCoIIhhBAiCAnrAIQ4Cr1Rj+KqYpTry2HkjdUT+Vt/VP+F53lIxVJopBq4y9whFUsZpSWEPSoY4tJMvAm5Fbm4UXYD18uu43rZddwou4ECbQFKqkpQoitBaVUpSqpKoDVqrV6+XCyHRqqBm8ytunTk7vBV+iJAFYBAdSAC1YEI0YQg1C0UcrFcgHdICDscz/M86xCECM1oMuJy8WWcLziP8wXnkVaUhmul15BVngW9Sc86Hjhw8FX6Itw9HM29miPGOwYx3jFo5tkMMrGMdTxCGoQKhjgdE2/C+YLzOJt3FqkFqTiffx4Xiy5CZ9SxjmY1CSdBhEcEWni3QAvvFmjv1x6tfVtDKqJdb8T+UcEQh8fzPC4WXcSRrCM4kn0Ex3KOobSqlHUswSglSnT074iugV3RJaAL2vi2gUREe7uJ/aGCIQ4ptyIXu6/uxuGswziWcwwF2gLWkZhRSVTo6N8RPYJ6YFCTQQhzD2MdiRAAVDDEgVwvu47EK4nYcWUHTt88bT5yi1hq5tUMcU3iMKjJIMR4x7COQ1wYFQyxa+nF6eZSSS1IZR3H4YS5hWFQk0GIC49De7/2rOMQF0MFQ+xOsa4YW9K3YOPFjVQqjSjCPQKPNHsEo6NGw0fpwzoOcQFUMMRuHM0+ip8u/ISdmTtRZapiHcdpSUQS9Avth0eaPYLY4FiIRWLWkYiTooIhTJVWlWJT2ib89NdPSC9OZx3H5fir/DEmegzGx4yHn8qPdRziZKhgCBM3K25i7bm1+Omvn1CuL2cdx+XJRDKMjBqJ+NbxaOrRlHUc4iSoYIhNZRRnYE3KGvzv0v9oN5gd4sBhQNgATGk7hQ4KIA+MCobYREpeCladXYWdmTth4k2s45B66OTfCVPbTkXf0L6soxAHRQVDBJVWmIZlJ5Zhz7U9rKOQBmrv1x4vdX4JnQI6sY5CHAwVDBFEbkUuvkj+Ar+m/Xr70vbEofUP7Y85neYg2iuadRTiIKhgSKMqqyrDf87+B9+nfo9KQyXrOKSRiTkxRkWNwswOMxGoDmQdh9g5KhjSKIwmI9ZfWI+vTn/l0tcFcxVysRzPtHoGz7V7DkqJknUcYqeoYMgDS8lPwTt/vkNn3bugEE0I3uz+JvqE9mEdhdghKhjSYBX6Cnx+8nOsO7+Oxllc3EPhD+G1bq/RyZrEAhUMaZDdV3fjvcPvIbs8m3UUYifcpG5I6JSAJ2KegIgTsY5D7AAVDLHKzYqbeP/I+9hxZQfrKMROtfNth0WxixDlGcU6CmGMCobU267MXVjw5wIU6gpZRyF2Ti6WY06nOZjQcgI4jmMdhzBCBUPqpDVo8fGxj7H+wnrWUYiD6RHUA4tjFyNAHcA6CmGACobc11+Ff+HVva8irSiNdRTioDzlnng39l30D+vPOgqxMSoYck8/pP6Afx3/F3RGHesoxAk81eIpzOsyDzKxjHUUYiNUMKSGcn053tj3BpKuJrGOQpxMW9+2WDZgGfxV/qyjEBuggiEWrpZcxexds2mXGBGMn9IP/xrwL7odgAuggiFmh7IOYf6e+SjWFbOOQpycTCTDWz3ewthmY1lHIQKigiEAqsdbPj76MQy8gXUU4kKeavEUXu76MiQiCesoRABUMC5Ob9TjvcPv4ZeLv7COQlxUt8BuWNpvKTwVnqyjkEZGBePCyvXlmJ00G0eyj7COQlxcpEckvhr8Fd0CwMlQwbioQm0hpidOR0p+CusohAAAgtRB+GrwV2jq0ZR1FNJIqGBcUHZ5Np7b8RzSi9NZRyHEgpfcCyvjVqK1b2vWUUgjoEueupiM4gxM/GMilQuxS4W6QkzdPhWHsw6zjkIaARWMC0nNT8WkrZOQVZ7FOgoh91SuL8eMxBlIvJLIOgp5QLSLzEWcuXkGz+94HqX6UtZRCKkXMSfGh30/xEMRD7GOQhqICsYFXCi4gCnbpqCkqoR1FEKsIhFJ8OmAT9E3tC/rKKQBaBeZk0svTsdzO56jciEOyWAy4KXdL+FQ1iHWUUgDUME4setl1/Hs9mdRoC1gHYWQBtMZdZidNBsnc0+yjkKsRAXjpHIrcjFt2zTkVOSwjkLIA6s0VGJG4gyk5NF5W46ECsYJFWoL8ez2Z3Gt7BrrKIQ0mjJ9GZ5PfB6Xii6xjkLqiQrGyeiMOsxKmoXLxZdZRyGk0RXrijFz50za7esgqGCcCM/zeGPfGzh98zTrKIQI5nrZdcxJmoMqYxXrKKQOVDBOZEXyV9h+ZTvrGIQILvlmMt4+8DbrGKQOVDBOYuPJa/hxpy/CNHShQOIatqRvwcpTK1nHIPdBBeMEkq8W4bVfzuB6nhJXz0xFS4+urCMRYhMrklfgj/Q/WMcg90Bn8ju4nBItRi/fj5wSnXmaWMQjttufOFn8P4bJCLENuViOtcPWopVPK9ZRyF1oC8aBGU08EtadtCiX6ukc9h6KRTvFFEg4uhUtcW46ow7z98xHWVUZ6yjkLlQwDuzTnRdxJP3eh2seONkcTfSzoZG62TAVIbZ3tfQqFh5cyDoGuQsVjIM6eCkfy5Mu1jnfqYv+kObOQZAq1AapCGFnW8Y2bLiwgXUMcgcag3FAheVVGPbpPmSXaOv9Gg+NHs3a/BcXiul6TsR5ycVy/DD8B8R4x7COQkBbMA5p/k+nrCoXACguk+LU0cfQwXOIQKkIYe/WeEyFvoJ1FAIqGIezan86dp7PbdBr9UYR9h0cgPaqiRBx9E9PnFNGSQbeO/we6xgEVDAOJTWrBB/8cf6Bl7P/eCtEGhOgkqgbIRUh9mfzpc3Ye20v6xgujwrGQZhMPF775TSqjKZGWd7JC0FQ589FgDKoUZZHiL1ZdHARHbrMGBWMg1j9ZwZOXStu1GVevuGGgovTEe3eplGXS4g9yKnIwdLjS1nHcGlUMA7gWmEFlm6/IMiyC0plSDnxFNp7DhRk+YSw9Mtfv+BI1hHWMVwWFYwDeGvTWVRUGQVbfpVehP0HH0IHzZPgwAm2HkJsjQePBX8uQKWhkmmOhQsXokOHDla9JiIiAsuWLRMkj61Qwdi5X5OvY/eFmzZZ176j7RHDzYRCrLDJ+gixhWtl1/DZic8afbkHDx6EWCzGiBEjGn3ZteE4Dps2bapzvoULF4LjuBqPFi1aCB/yLlQwdqyoogrv/nbOpus8ei4UnsVz4avwt+l6CRHSuvPrcLGw7itfWGPVqlVISEjA3r17cePGjUZd9oNq3bo1srKyLB779++3eQ4qGDv24bYLyCuz/V37Ll71RFn6DES62f43HkKEYOSN+OjoR422vLKyMqxfvx7Tp0/HiBEjsGbNGovn//nPfyIgIABubm6YOnUqtFrLE6P79++PuXPnWkwbM2YM4uPja11fREQEAGDs2LHgOM789b1IJBIEBgZaPHx9fc3PZ2VlYcSIEVAqlWjatCl+/PHHGrvkMjMz8fDDD0Oj0cDd3R2PP/44cnJy7rveu1HB2KmLOaVYf/Qqs/XfLFLgQvIzaOvZh1kGQhrTwayDjXZuzIYNG9CiRQvExMRgwoQJ+M9//oNbV93asGEDFi5ciCVLluDYsWMICgrCihUrHmh9R48eBQCsXr0aWVlZ5q8bauLEibhx4wZ2796NX375BV9//TVyc2+fwG0ymfDwww+joKAAe/bswY4dO3D58mU88cQTVq2HCsZO/fOP8zCa2F4mTlslxsFDw9HRbRzTHIQ0lo+PfQyDyfDAy1m1ahUmTJgAABg6dCiKi4uxZ88eAMCyZcswdepUTJ06FTExMVi8eDFatXqwe9X4+fkBADw9PREYGGj++l7OnDkDjUZj8XjhhRcAAOfPn0diYiL+/e9/o3v37ujUqRO++eYbVFbePhBi586dOHPmDH788Ud07twZ3bt3x9q1a7Fnzx6ryo0Kxg4dupzf4MvBNDae57D3SBe0krwAuVjOOg4hDyS9OB3rL6x/oGVcuHABR44cwZNPPgmgenfUE088gVWrVgEAUlNT0b17d4vX9OzZ84HWWZvMzEyLAlmyZIn5uZiYGCQnJ1s8Fi1aZM4vkUjQqVMn8/zR0dHw8vIyf52amoqwsDCEhYWZp7Vq1Qqenp5ITU2td0a6G5Wd4Xke72+p/z+grRw+E4GWEXNQ7rkKhbp81nEIabCVp1ZiZORIeMg9GvT6VatWwWAwIDg42DyN53nI5XIsX768XssQiUS4+0L2er3eqhzBwcFITk42f+3t7W3+u0wmQ3R0tFXLEwJtwdiZ/53OavQz9htLaoY39JmzEK5h/x+XkIYq1hXjq9NfNei1BoMBa9euxdKlSy22Dk6dOoXg4GCsW7cOLVu2xOHDhy1ed+jQIYuv/fz8kJWVZf7aaDTi7Nmz9123VCqF0Xj7fDiJRILo6Gjz486CuZ+YmBgYDAacPHn71h1paWkoLCw0f92yZUtcvXoVV6/eHgc+d+4cioqKrNrdR1swdqTKYMLH24Q5Y7+xZBUoUVIej7adfkdK0UHWcQhpkA0XNmBy68nwU91/LONuv/32GwoLCzF16lR4eFhuAT366KNYtWoV5s+fj/j4eHTp0gWxsbH44YcfkJKSgsjISPO8AwcOxEsvvYTff/8dUVFR+OSTT1BUVHTfdUdERGDnzp2IjY2FXC632KV1N4PBgOzsbItpHMchICAALVq0QFxcHJ577jmsXLkSUqkU8+bNg1KpBMdVn2gdFxeHtm3b4umnn8ayZctgMBgwY8YM9OvXD126dKn394u2YOzI/x3NRGaB/d/HolwnwZFDo9DRYwzrKIQ0iM6ow6qzq6x+3apVqxAXF1ejXIDqgjl27BhatmyJt99+G6+88go6d+6MK1euYPr06RbzTpkyBZMmTcLEiRPRr18/REZGYsCAAfdd99KlS7Fjxw6EhYWhY8eO9503JSUFQUFBFo/w8HDz82vXrkVAQAD69u2LsWPH4tlnn4WbmxsUiuqTrDmOw6+//govLy/07dsXcXFxiIyMxPr11o1f0R0t7YTBaEL/j3fjWiHbS1pYq2f7SzhvWN0oR+YQYktysRxbHtkCfxWdVHzt2jWEhYUhMTERgwYNarTl0haMnfjtdJbDlQsAHDwVhZDKOfCQNWzAlBBWdEYd1qSsYR2DiaSkJGzevBnp6en4888/MX78eERERKBv376Nuh4qGDvx5Z5LrCM02NnLfkDWbISow+uemRA78vNfP6NIW8Q6hs3p9Xq88cYbaN26NcaOHQs/Pz/s3r0bUqm0UddDu8jswK4LuZi8+sHOzLUHbio9WrTbhPPFx1lHIaTeXmj/AmZ2mMk6hlOiLRg7sHK342693Km0QoqTRx5FRw/bXF2WkMbwY+qPzC/n76yoYBg7kVmII+kFrGM0GoNJhL2H+qC9cjLEnJh1HELqVFJVgi2Xt7CO4ZSoYBj7yoHHXu5n/4kYhBtmQyPVsI5CSJ0e9PIxpHZUMAzllmiRmGof1xwTwqm/AiC/ORdBqlDWUQi5r9SCVJy+eZp1DKdDBcPQT8evMb9istAysjTIvfAcmnu0Zx2FkPuirZjGRwXDCM/z+OkYu/u92FJRmQynjz6BDp4PsY5CyD1ty9jmkocsC4kKhpFDlwuQkW//l4VpLHqjCPsODkQH9QSIOPpvR+yPzqjDprRNrGM4FfqkM7L+aCbrCEzsO9YG0fwsqCQq1lEIqeGnv35iHcGpUMEwUFypxx9ns+ue0UkdTw2GpmAu/JWBrKMQYiGzNBNnbp5hHcNpUMEw8GvydegMJtYxmLp03R1FaTMQ5d6adRRCLPyR8QfrCE6DCoaBjSevs45gF/JLZEg98RTaed7/MuWE2NL2jO017jZJGoYKxsZySrRIvlrEOobd0OnFOHBwCDpqngAHjnUcQpBTkYMTuSdYx3AKVDA2tv1cDuiXo5r2Hu2IFqIZkIsVrKMQgj/SaTdZY6CrKdvYM6sOY9/FPNYx7FZMk0LofFYhX0vfI1vQF+qRvSEbZafLYKoyQRYgQ+jUUCibKgEAORtzUHy4GPoCPTgJB2WEEgGPBkAVde+jAPOT8lGQVAB9nh4AIA+Rw/9hf7i1czPPk7UuC0X7i8DJOQSOC4RnL0/zc8VHilF0oAjhL7K7/YO3whtJjyVBLKLr6T0ICesArqS4Uo9Dl/NZx7BrFzK94F8yCxHNf0RG6V+s4zg1Y7kRlxdfhrqlGuHzwiFxk0CXo4NIfXvHhjxQjuBngiHzk8GkNyF/Wz4yPs5A8w+aQ+Je+48PqZcUgY8FQhYgAwAU7S9C5qeZiFoUBUWIAiUnS1B8sBgR8yOgy9Hh+qrr0LTVQOImgbHCiJxfchDxSoQtvgX3VKAtwNGco+gR1INpDkdHu8hsaNf5XOiNtMFYl9wiBdKSJ6GNZ2/WUZzazd9vQuojRei0UKgiVZD5yeDWxg1yf7l5Hs+entC01kDmL4MiRIHAJwNhqjRBe017z+W6d3SHW3s3yAPlkAfKETAuACKFCBVp1ScW67J0ULdQQ9lUCc8enhApRai6WQUAyN6QDe+B3pD5yIR98/Ww5+oe1hEcHhWMDW1Lcd1zX6xVWSXGoUMj0NH9UdZRnFZpcimUEUpkLs9EakIq0v6RhoLd9751hMlgQuHuQoiUIijC6jdWxpt4FB0qgklngiq6ereaIkyByoxKGMuNqMyoBF/FQx4gR/lf5dBe0cJnsE+jvL8HdfDGQdYRHB7tIrMRrd6IPX/dZB3DofA8h72Hu6JHO19cNK5GlamKdSSnUpVbhYKkAvgM9YHfKD9Uplci64cscBIOXr29zPOVJJfg2sprMFWZIPGQIOLlCEjc7v+jQ3tVi8uLL8OkN0EkF6FJQhMoQqpLya2tGyp6VuDSO5fAyTiEPhsKTs7hxtobCJ0WioKkAuQn5kOikSB4crD5dbZ2qfgScspzEKAOYLJ+Z0CD/DZy8FI+nvz3IdYxHFarpvko9fgGRbpC1lGcRsrUFCiaKhD1VpR52o3vb6AyvRJRb9+eZtKZoC/Sw1hqRMGeApSnliPqH1H3HIMBqrd29Pl6mCpNKD5ajMK9hWj6WtN7lkXuplwYK4zw6uOFjI8yEL04GqWnSpGfmI/od6Ib701baVGvRRjbbCyz9Ts62kVmI4fTaXD/QZxL94HxWgKaaCJZR3EaEk8JFMGWP/DlwXLo8/UW00RyEeQBcqiiVQidGgpOzKFw7/2LXiSpfo0yQonAxwKhCFMgf0ftnwHdDR2KDhbB/xF/lJ8vhypGBYm7BB7dPKC9ooWx0vhgb/QB0G6yB0MFYyOHLzvPbZFZuZGnwpUzU9DKszvrKE5B1UwFXbbOYlpVdhWkvtL7vo438TDprbzUEQ/w+po7S3iex/VvryNwfCDECjF4Ew/+7wNheMPf8zO8qtKhrEN0Vv8DEKRgOI7Dpk2b7vn87t27wXEcioqKAABr1qyBp6fnfZe5cOFCdOjQodEy2lKVwYSTV2nXTmMo00pw7PAYdHAfxTqKw/N5yAcVlyqQ+79c6HKqtyIKdhfAZ2D1ILtJZ0L2z9moSKtAVV4VKjMqcW3VNRgKDfDo5mFeTvoH6chPvL11kv1TNsovlKPqZhW0V7XVX58vh2dPzxoZCvcUQuImgXtHdwDVpVeeWo6KtArkbc+DPFgOsZrduSiFukKcKzjHbP2OrkEFk52djYSEBERGRkIulyMsLAyjRo3Czp076/X6Xr16ISsrCx4eHnXPbIX4+HhwHFfjMXTo0EZdj7VOXyuC1trf+Mg9GU0c9h2ORVv5VEhEdJxKQ6kiVWiS0ATFh4uR9mYabm6+iaCngm6f9MgBVVlVyFyeiYuvXcSVZVdgLDOi6RuWYylVuVUwlBrMXxtKDLj29TVcfP0i0j9MR2V6JSLmRUDTRmOxfkOxATf/dxNBTwdZZPId6osr/7qCkiMlCJkWIuj3oD6OZR9jHcFhWf3pzMjIQGxsLDw9PfHRRx+hbdu20Ov12LZtG2bOnInz58/XuQyZTIbAQGEu1T506FCsXr3aYppcLr/H3LZxOJ12jwnhz+RmaBs1B/nqf6NUX8I6jkNy7+AO9w7utT4nklUf/VWXmKUxFl+HTg2t17olHpIarwUA/4f94f+wf72WYQtn8ujy/Q1l9RbMjBkzwHEcjhw5gkcffRTNmzdH69at8dJLL+HQodtHSeXl5WHs2LFQqVRo1qwZNm/ebH7u7l1ktfnnP/+JgIAAuLm5YerUqdBq731i153kcjkCAwMtHl5etw+5PH/+PHr37g2FQoFWrVohMTGxxi69M2fOYODAgVAqlfDx8cFzzz2HsrKy+n+T7kJn7wvnzCU/SHLmIFhd9w9CQhribN5Z1hEcllUFU1BQgK1bt2LmzJlQq9U1nr9zHOWdd97B448/jtOnT2P48OF4+umnUVBQv9/kN2zYgIULF2LJkiU4duwYgoKCsGLFCmui1spoNGLMmDFQqVQ4fPgwvv76a7z55psW85SXl2PIkCHw8vLC0aNH8dNPPyExMRGzZs1q2DpNPE5cofEXIWXmqJGV+ixaeHRiHYU4oetl11Gopc9wQ1hVMGlpaeB5Hi1atKhz3vj4eDz55JOIjo7GkiVLUFZWhiNHjtRrPcuWLcPUqVMxdepUxMTEYPHixWjVqlW9Xvvbb79Bo9FYPJYsWQIA2LFjBy5duoS1a9eiffv26N27N9577z2L1//444/QarVYu3Yt2rRpg4EDB2L58uX47rvvkJOTU68Md0rPK0N5FbvDLF1FSbkUJ4+MQwdPtuNtxDnRVkzDWFUw1hyu165dO/Pf1Wo13N3dkZubW6/Xpqamont3y0NRe/bsaf77vn37LArkhx9+MD83YMAAJCcnWzxeeOEFAMCFCxcQFhZmMf7TrVu3Gutu3769xRZabGwsTCYTLly4UK/8d0q5QWMDtmIwibDvYH+0U06EmKOr4JLGczafCqYhrBrkb9asGTiOq9dAvlRqeSw9x3EwmRrnSKouXbogOTnZ/HVAwO1LOajVakRHszvz927nsqhgbO3AiVbo0DwB2YpvUK5v+NgZIbek5KWwjuCQrNqC8fb2xpAhQ/DFF1+gvLy8xvP3G7S3RsuWLXH48GGLaXceQKBUKhEdHW1+uLm53b2IWsXExODq1asWu7qOHj1aY92nTp2yeH8HDhyASCRCTEzNI17qkppVavVryINL/isQyptzEKgKZh2FOAHaRdYwVh9F9sUXX8BoNKJbt2745ZdfcPHiRaSmpuKzzz6z2I31IObMmYP//Oc/WL16Nf766y8sWLAAKSn1+w1Cp9MhOzvb4pGXV33zqsGDByMqKgqTJk3C6dOnceDAAbz11lsAqrewAODpp5+GQqHApEmTcPbsWezatQsJCQl45plnLLaU6uuvbCoYVtKz3JD31wto5t6WdRTi4PK1+Sipor0R1rK6YCIjI3HixAkMGDAA8+bNQ5s2bTB48GDs3LkTK1eubJRQTzzxBN5++2288sor6Ny5M65cuYLp06fX67Vbt25FUFCQxaN37+r7iojFYmzatAllZWXo2rUrpk2bZj6KTKGoPnFMpVJh27ZtKCgoQNeuXTFu3DgMGjQIy5cvt/p9lGj1yC6p3+HVRBiFpTKcPfYk2nvGsY5CHFxmSSbrCA7H5a+mfODAAfTu3RtpaWmIioqq+wVWOH6lEI+u/LNRl1kX7dWzKDn8C6pyLsFYVgC/sW9C1dxyy1KfdxWFe1ZDm3kW4I2Q+jSB39jXIXGv/eS27B9fg+5qzV0Eysgu8H9sIQCg+PB/UXLkFwCAR/dH4d7tEfN8uhsXULB9BQInfgKO4S1o+3Q5g1PlP4KHS/+XJw30zz7/xIjIEaxjOBSXu87Gxo0bodFo0KxZM6SlpWHOnDmIjY1t9HIBgEs3bT/AzFdpIfWPhKbdYNzcuKTG8/rCLGT/8Ao07QbDs/fT4GQq6PMywYnvfQdBv7FvAsbblwIxVpYga3UCVC2qtwyrctNRvP8H+I37B8DzuPnLIiiadoLMLwK8yYj8bV/AZ+gspuUCAPuOtUWXlrOQKfkGlYZKplmI46EtGOu5XMGUlpbi1VdfRWZmJnx9fREXF4elS5cKsq5rhbb/IaaM6gJlVJd7Pl+0dy2UUV3gNWCKeZrUK+ie8wOAWGl5EEV56l5wUjlUMdUFo8+/BqlfBJTh7auX5xcBff41yPwiUHL4FyjCWkMe1Lyhb6lRHUsNQXToXKj9VyGvsn6HzRMCAFdKr7CO4HBcrmAmTpyIiRMn2mRd2cX29Vsyz5tQefkY3Ls9gpz1b6Mq9zIkHgHw6PFYjd1o91N2ejvULftCJKset5L5RcBQeB2GklyABwwF1yHzDYe+MAtlZxIRNGmZQO+oYdKuecCvdAYiY9bjcmkq6zjEQVwtuco6gsOh+8EIKKvYvgb4TeXF4KsqUXL4ZygjOyPg8Xehat4TNzcugTazfhf00924AH3eFWjaPWSeJvUNg2ffichZ/zZyNrwNz36TIPUNQ8G25fDqPxmV6SdwY9UM3Fg9G9paxnJYuFmswIXkCWjr2Y91FOIgaAvGei63BWNL2XZWMDxffaKrMroH3LuOAQDIAiKhu56K0uQ/oGhS9+G8Zad3QOoXAXmw5TlBbh2Hw63j8NvzndkJTqaEPKQFrv/7BQRN/ATG0nzkbf4QIc+vAie5/02tbEFbJcbBQ0PRp2sATpZuYB2H2LliXTHKqsqgkWnqnpkAoC0YQdlbwYhV7oBIDKlvmMV0qU8YjCU363y9qUqL8tS90LQbfN/5jBXFKD7wI7zjXoDuxl+QegdD6h0CRXg78EYD9IXXH+h9NCae57D3SCe0FM+AXMz2tg7E/tFFL61DBSOQMp0BpTpD3TPaECeWQh7YDIYCyx/w+oLrEN/jEOU7VVzYD96oh7r1gPvOV5j0Ddy6joHE3RfgjeCNd1zs02QEGumSQY3pyNkm8CmdC2+5D+soxI4V6qhgrEEFIxBWA/ymqkpU5VxGVc5lAIChOAdVOZerB+ABuHd/BOWp+1CavBX6whsoOf4/VKYdgVun27u38n5bisI9a2osu+z0dqia9YBYWfsNqgCgMv0k9AXX4dap+nwBWWBzGAquofLSMZQmbwVEYki82d+lsDYXrnhBd2UWItyasY5C7BRtwViHxmAEklOiY7LequyLyFn3hvnrwqRvAADqNoPgO+JFqJr3gs+QGSg+9BMKd34NiXcI/Ma+AUVoa/NrDCU3Ac7ydw99/jXorp2D/+Pv3nPdJr0OBYlfwm/0q+D+fr3E3Rdecc8j749l4MRS+Ix4ESKp/e6Kyi5UouRUPNp1+h0pRbY9SZbYvwIt3Z3WGi5/Jr9Qtp7Nwgvfn2AdgzSQiDOhd7ejOFmykXUUYkde6vwSJreZzDqGw6BdZAKpoJuMOTQTL8Lew93RRvYcpCL2R7wR+0C7yKxDBSMQKhjncPBUJIIr58JD5sk6CrEDtIvMOlQwAqmkgnEaZy/7gL8xG2GapqyjEMa0Rvs69cDeUcEIhLZgnMv1mypcPTMFLT26so5CGDKY7OvUA3tHBSOQCj39R3Q2pVopjh95BB09RrKOQhihgrEOFYxAaBeZczKaOOw91BvtFJMh4egof1dDBWMdKhiBUME4twMnY9BEPxsaqVvdMxOnYeCpYKxBBSMQjmOdgAjt1EV/SHPnIEgVyjoKsRHagrEOFYxApGL61jozjuMxKfg6Vrttw2SxN+s4xEaoYKxDO5EFQgXjnPp6FyHB5xg6Fu+ApOAqUACc0wxkHYvYiNFEu76tQQUjEKmY9pE5i0iVFvOCz6C/bhfUN5OBCsvn5fRbrcuQiWWsIzgUKhiB0BaMY1NLjJgbehFjRPvhm70P3DX9PedVGKhgXAXdM8g6VDACkVDBOByO4/FMUBYmqg8iKjcRXHZxvV4nN927fIhzoYKxDhWMQGS0i8xh9PYuxmyfY+hUvAOSgkzAystNyQ16+iS5CIVEwTqCQ6GPhUBoF5l9i1BqMT/kLAbokmodV7GGwlBFnyQXoZaqWUdwKPSxEIinii7xbm/UEiPmhKZhjGg//LL33ndcxRpyA5ubyxHbo4KxDhWMQHw1tK/WHnAcjwlBNzBRfRjRuTvqPa5iDYWhCgDtEnUFGqmGdQSHQgUjEB8qGKZ6excjwec4Ohdvb9C4ijXkVZUAVMKtgNgNd7k76wgOhQpGID5qOl7e1iKUWswLScEAXRI0N08+0LiKNRQGLahgXEOAKoB1BIdCBSMQPzfagrEFtdiE2aFpGCve16jjKtao3oIhrsBf5c86gkOhghGIQiqGWiZGOV1VWRBPB91AvOYwom7ugCiniGkWMW+ERCSh61S5ACoY61DBCMjXTY7yfBvtp3EBsV7FSPA9js7FOyAtvAIUsk50m0IsRxkVjFMTc2L4Kf1Yx3AoVDAC8tPIcYUK5oE0UWoxPyQFA3W7oLl5ArjKOlHt5CIZylDOOgYRkI/CB2KRmHUMh0IFI6Am3iocu2JHv2Y7CLXYhITQNIwV74d/zl5w16pYR6qTgi6C6PRo95j1qGAEFOlHJ2VZ46mgLMRrDiH6ZiJEOY5VzHIRnVjr7KhgrEcFI6AoPzopqy49vYoxx/cEOpfsgLQww67GVawh52jXibML9whnHcHhUMEIKJIKplahCh1eDknBIP0uaHKP2+24ijXkHH2UnF0zz2asIzgc+lQIKMJXBREHmHjWSdhTi02YFXoJY8X7EZCzB9x1+x9XsYacBn+dXpRnFOsIDocKRkByiRihXipkFrjukWTjg7IwWXMIzRxwXMUaClDBODMRJ0KkRyTrGA6HCkZgkX5qlyuY7p4lmOt3Al1KdkBamO6w4yrWkHN0ewZnFqoJpXvBNAAVjMCi/DTYfeEm6xiCC1XoMD80BYOqdsMt95hTjKtYQwEqGGdGu8cahgpGYO1CPVhHEIxabMLM0Et4RLIfAdl7wV1z3fui0JXnnFu0ZzTrCA6JCkZgnZp4sY7Q6J4IysYUzWE0y9sBUY6A18F3IHKe7gfjzJp7N2cdwSFRwQgszFsFX40ceWWO/dt9d88SzPE7ga4liZAWXnaJcRVrKECHCjqzjn4dWUdwSFQwNtCpiSe2n8thHcNqIQod5oWcw2D9LpccV7GGnI5Fd1ohmhAEqOk+MA1BBWMDHZt4OUzBKMVGzAxNxzjJfgRk7wF33bG3vGxFwVPBOKuO/rT10lBUMDbQqYkn6wh1eiwwG1PdDqN5XiJEOfms4zgcucnEOgIRCBVMw1HB2EC7UE9IRBwMdrYbpZtnCeb4nUTX0h2QFV0GilgnclwKE91Yzll18u/EOoLDooKxAaVMjNbB7jh1rZh1FItxFU3ucXBX7av0HJWcCsYpecg96ByYB0AFYyN9m/sxKxil2IgZIRkYJ92PwOzdNK4iALmR7mbpjDr6dQTH0SHoDUUFYyP9Y/zxeVKaTdc5LjAH09wOVY+r5NK4ipAURj3rCEQAvUJ6sY7g0KhgbKRjmCe81TIUlAt7FeEuHqWY638C3UoTISu6ROMqNiI36kFXi3E+/UP7s47g0KhgbEQk4tCvuR82nrze6MsOUlRhfkgKBuurrwNG4yq2JzdUAXTXZKfS3Ks5gjRBrGM4NCoYG+of03gFoxQbMSM0A49KDiAoZze469pGWS5pGLlBTwXjZPqF9mMdweFRwdhQ/+b+EIs4GB/gcOVHA3Iwzf0wYvJ20PkqdkRhpAMnnE3/sP6sIzg8Khgb8lBJ0THME8euWHchr04epXjR/yS6l+6oHldhf7QzuYtcrwN9nJyHr9IXbX3bso7h8OgTYWODWwXUq2CCFFWYF3wODxl3wy3nKI2r2DmFvhKAG+sYpJH0De1Lhyc3AioYGxvZPhj/3HoetV26Si4yYUZoBh6T7q8eV7lB4yqOQq7XggrGecQ1iWMdwSlQwdhYiKcSXcK9cDTj9lbMIwE5mOZ+BC3ydkCUm8cwHWmo6oIhzsBX6YtewXT+S2OggmFgdIcQGAuvYq7/SfQo3QFZURqNqzg4DjzkYjl0NNjv8EY0HQGxSMw6hlOggmFgfDtPTEicDe4q/dbrTORiGRWMExgdPZp1BKdB5x4zIFV5gGs+hHUM0sgUIjoRxtG18G6B5l50e+TGQgXDSocJrBOQRiYXSVlHIA9odBRtvTQmKhhWogcBbnQZCmdCBePYJJwEw5sOZx3DqVDBsCISA+3Hs05BGpFCREOajqxfWD/4KH1Yx3AqVDAsdX0WoN96nYaco4JxZBNa0m7rxkYFw5JHCNDmUdYpSCNRcHRoq6Nq6d0SXQK7sI7hdKhgWOuVwDoBaSRyjj5OjuqZVs+wjuCU6BPBWmAbIGog6xSkEcjp4+SQ/JX+GNp0KOsYTok+Efag12zWCUgjkIMujuiInmz5JKQ0FioIKhh7EDUACGzHOgV5QHLWAYjVlBIlHmv+GOsYTosKxl7QVozDU/C0BeNoxkSPgYfcg3UMp0UFYy9ajwU8wlinIA9AXts9GIjdUogVeLbts6xjODUqGHshltARZQ5OQQXjUJ5s8ST8VH6sYzg1Khh70nky4NWUdYoGe3+fDl3/XQa390vg/1EpxvxfBS7kGS3m0Rp4zPy9Ej4flkKzpASPbqhATpnpvsvleR7/2KVF0NJSKN8rQdzaclzMv71cnYHHMxsr4f5+CZp/XobEywaL1390QIeELZWN90bvQW66//sg9kMj1WBKmymsYzg9Khh7IpEBcQtYp2iwPVcMmNlVhkNT1djxjAp6E/DQ9xUor7r9m/2LW7X4318G/PSYEnvi1bhRyuORDff/4f/hgSp8drgKX45Q4PA0NdQyDkO+r4DWUL3cr4/rcfyGEQenqvFcZyme+qUS/N9bE+mFJvz7hB7vDVII98b/puCpYBzFM62egafCk3UMp0cFY29ajwVCu7FO0SBbJ6gR30GG1v5itA8UY83DCmQW8zieVb21UazlseqkHp8MUWBgUwk6B4ux+mEF/rxqxKFrhlqXyfM8lh2uwlt95Xi4hRTtAsRYO0aJG6U8Np2vfk1qnhGjYyRo7S/GzK4y3KzgkVdRXTDTf6/EB3FyuMuFH4CXG411z0SY85B7YGKriaxjuAQqGHs05D3WCRpF8d/33vJWVv9wP55lhN4ExEXevmZXC18xmnhwOHi19h/O6UU8sst4i9d4KDh0DxWbX9M+QIz9mUZU6nlsu2RAkIaDr4rDD6f1UEg4jG1pm3Mc5KbaS5LYlyltpkAj07CO4RLo6nz2KKwb0Oph4NyvrJM0mInnMXerFrFhYrTxr75GV3YZD5kY8FRYbk0EqDlkl9U+QJ799/hMgLqW15RXPzeloxSnc4xotaIMvioOGx5TolAL/GO3FrsnqfFWkhb/d1aPKG8R/jNaiRB3YX6vUtAWjN0LUgfhqRZPsY7hMqhg7FXcQuDCH4CxinWSBpn5uxZnc43YP0Ut+LqkYg5fjFBaTJv8ayVmd5PhZLYRm84bcOoFDT48oMPsrVr88rhKkBxyB/23ciXzusyDQiL8eBypRrvI7JV3JNB1GusUDTJrSyV+u2jArklqhN6xtRCo4VBlBIq0llsrOeU8AjW1j5EEakTmeWq8Rl37f99d6Qak5Boxq5sMuzOMGN5MArWMw+OtpdidIdxWhsKgF2zZ5MF1C+yGIRF0q3JbooKxZ31fBhzoSBee5zFrSyU2njcgaaIKTb0s/3t1DhJDKgJ23nEY8YU8IzKLefQMq/1S9009OQRqOIvXlOh4HL5mrPU1WgOPmVu0+GqkEmIRB6MJ0P/dKXoTYDQJd64KbcHYLwknwWvdXmMdw+VQwdgzlTcweBHrFPU2c4sW35/W48dHlHCTc8guMyG7zIRKffUPdQ8Fh6kdpXhpuxa70g04fsOIyb9q0TNUjB6hdwz8Ly/DxtTqrQGO4zC3uwyL9+mw+YIeZ3KMmLixEsFuHMa0qLmH9909OgxvJkHHoOryiW0ixn/P63E6x4jlR6oQ20S4vcIKAxWMvXqm1TNo5tWMdQyXQ2Mw9q7zJODcJuBSEuskdVp5rLoU+n9bYTF99cMKxHeQAQD+NVQB0TYtHt1QAZ0RGBIlwYoRlvvEL+SbUKy7vaXxSqwM5Xoez/1PiyItj95NxNg6QQWFxHK32tlcIzacMyD5+dvjPuNaSbA7Q4I+q8sR4yPCj48KM/4CAHKDTrBlk4YLUgfhhfYvsI7hkjiep+tb2L2iq8CKnkBVKesk5D7yNP4Y4EcDyPbm0wGfYmATuucSC7SLzBF4hgEPOc6uMlelMGhZRyB3GRE5gsqFISoYR9F5MtC0H+sU5D7kVcJf74zUX4AqAG90f4N1DJdGBeMoOA4Y/TlAZyDbLalJDzFX+9FwxLY4cHg39l24y9xZR3FpVDCOxCu8+gRMYrdkYhnrCATVl+LvGdyTdQyXRwXjaLpOAyL6sE5B7kEhooJhLcI9Ai92fpF1DAIqGMfDccDYLwGVD+skpBZy2oJhSsJJ8H6f9+lyMHaCCsYReYQC4/4D0P5+u6MQ2ebKzaR2CZ0S0Ma3DesY5G9UMI4qsj8w8E3WKchd5FQwzAwOH0x3qbQzVDCOrPdLQIuRrFOQO8g5ujgGC009muLd2HdZxyB3oYJxZBwHjFkJeEexTkL+pqDdljanlqqxbMAyqKXC3xqCWIcKxtEp3IEnvgfow2UX5FQwNvdu7LuI9IhkHYPUggrGGQS0AkZ/xjoFAaDg6CNlS5PbTMbg8MGsY5B7oE+Ds2g7Dugxk3UKlydH7TdOI42vf1h/zOk4h3UMch9UMM7kocVAm0dZp3BpVDC20c6vHT7q+xHEItolac+oYJyJSASM/QqIoqvHsqLgqWCEFuEegS8GfkEnUzoAKhhnI5ZWD/qHdGadxCXJQbdXEpKv0hcr41bC04FuJe7KqGCckUwNPP0z4BvDOonLUVC/CEYtVWPFoBUIdQtlHYXUExWMs1J5A8/8F3CnD6MtyU0m1hGcklQkxSf9PkFLn5asoxArUME4M49Q4JmNgNKbdRKXIeepYBqbVCTFJ/0/Qa+QXqyjECtRwTg7v+bVu8tkbqyTuAS5ycg6glORiWRYNmAZ+of1Zx2FNAAVjCsI7QxM2kxbMjZAu8gaz61y6Rval3UU0kBUMK4ipBMw+Q/ALZh1EqemMBpYR3AKcrEcnw38DH1C6eZ6jowKxpX4twCmbAW8mrJO4rTkVDAPTCFW4LOBnyE2JJZ1FPKAqGBcjVc4MGUb4N+adRKnpDDqWUdwaG4yN6yIW4FewTSg7wyoYFyRWwAw+XcgtBvrJE5HbqxiHcFhBauD8d2w79A1sCvrKKSRUMG4KqUXMHETEDmAdRKnotDTFkxDtPRuiR9G/IAoT7q3kTOhgnFlMjXw1Aag3XjWSZyG3KhjHcHh9AnpgzVD18BX6cs6CmlkVDCuTiIDHvkKGPoBIKLb/T4ohZ4Kxhrjmo/D5wM/h0qqYh2FCIDjeZ6unkSqZRwAfpoElN9kncRhFap90Nef7i5aFwknwYudX8TE1hNZRyECooIhloqvA+snADdOsE7ikCpkanQP8WEdw675K/3xcf+P0dG/I+soRGC0i4xY8gipPlem4wTWSRySQl/JOoJd6x7YHRtGbaBycRFUMKQmiRx4+AtgxFJAJGWdxqGIeBOk9D2rgQOHaW2n4avBX8FHSVt4roJGdcm9dZ0GBLYDfpkGFF1hncZhKMRy6E10uPItbjI3LOm9hC5Y6YJoC4bcX1g3YPoB2mVmBblIxjqC3YgNjsXG0RupXFwUbcGQusndqneZxYwA/jebjjKrg1xMu8jUUjXmd5mPcc3HsY5CGKItGFJ/LYYDMw4BrcawTmLXFC4+BtMtsBv+O/q/VC6EtmCIldS+wOPfAuc2A1vmA2U5rBPZHRnnmh8rpUSJuZ3m4skWT4LjONZxiB2gLRjSMK1GAzMPA+2fYp3E7ihc8IoIvUN645dRv+Cplk9RuRAz1/skkMaj9ALGrgS6TAa2vg5cP8Y6kV2Qc2LWEWwmWB2MV7q9gkFNBrGOQuwQFQx5cGHdgGmJwJmfgZ3vAMVXWSdiSsE5/44BhViB+DbxmNJmCpQSJes4xE45/yfhLvHx8eA4DhzHQSaTITo6GosWLYLBwPZOhPHx8RgzZkyd82VkZJjz3/04dOiQ8EHvheOAdo8Bs44CA98CZBp2WRiTO/nHaljEMGwesxkzO8ykciH35ZJbMEOHDsXq1auh0+mwZcsWzJw5E1KpFK+//rpVyzEajeA4DiKR7X+gJCYmonVry7tS+vjYwRnSUiXQ92Wg40Qg6V0g+QeAN7FOZVPOWjCxwbGY2WEm2vq1ZR2FOAjn/CTUQS6XIzAwEOHh4Zg+fTri4uKwefNmfPLJJ2jbti3UajXCwsIwY8YMlJWVmV+3Zs0aeHp6YvPmzWjVqhXkcjkyMzOh0+kwf/58hISEQK1Wo3v37ti9e3eN123btg0tW7aERqPB0KFDkZWVBQBYuHAhvv32W/z666/mrZE7X18bHx8fBAYGWjyk0tuHxy5evBj+/v5wc3PDtGnT8Nprr6FDhw7m500mExYtWoTQ0FDI5XJ06NABW7dubZTvL4Dqu2Y+vBx4fi8QPbjxlusAFKwDNLLuQd3x3bDv8OXgL6lciFVcsmDuplQqUVVVBZFIhM8++wwpKSn49ttvkZSUhFdeecVi3oqKCnzwwQf45ptvkJKSAn9/f8yaNQsHDx7E//3f/+H06dN47LHHMHToUFy8eNHidR9//DG+++477N27F5mZmZg/fz4AYP78+Xj88cfNpZOVlYVevRp+T/IffvgB7733Hj744AMcP34cTZo0wcqVKy3m+fTTT7F06VJ8/PHHOH36NIYMGYLRo0dbZG4UgW2BCT8DL+wH2j4GuMAAuJx3jqOougR0weohq/HNQ9+gg38H1nGIA3K5y/XHx8ejqKgImzZtAs/z2LlzJ0aOHImEhAR89NFHFvP+/PPPeOGFF5CXlwegektk8uTJSE5ORvv27QEAmZmZiIyMRGZmJoKDg82vjYuLQ7du3bBkyRLz69LS0hAVVX1L2BUrVmDRokXIzs6uket+MjIy0LRpUyiVyhq75m5tbfXo0QNdunTB8uXLzc/17t0bZWVlSE5OBgCEhIRg5syZeOONN8zzdOvWDV27dsUXX3xR32+n9QqvAAeXAye+AwzOeeXhzzqMwL+Lz7CO0WDdg7rj2bbPontQd9ZRiINzyTGY3377DRqNBnq9HiaTCU899RQWLlyIxMREvP/++zh//jxKSkpgMBig1WpRUVEBlar6jnsymQzt2rUzL+vMmTMwGo1o3ry5xTp0Op3FmIhKpTKXCwAEBQUhNzf3vjmHDRuGffv2AQDCw8ORkpJifm79+vVo2bJlra+7cOECZsyYYTGtW7duSEpKAgCUlJTgxo0biI2NtZgnNjYWp06dum+mB+YVDgz/COj3KnD4K+Dov4HKQmHXaWNyB/ydTSlRYmTkSDzV4ilEe0WzjkOchEsWzIABA7By5UrIZDIEBwdDIpEgIyMDI0eOxPTp0/Hee+/B29sb+/fvx9SpU1FVVWUuGKVSaXEiWVlZGcRiMY4fPw6x2HL3j0Zz+0iqO8dHAIDjONS18fjNN9+gsrKy1teHhYUhOtqBfxCofYGBbwK95wLHv60umoLLrFM1CoUDHdQQpA7C+Bbj8WizR+Eh92AdhzgZlywYtVpd44fz8ePHYTKZsHTpUvOupw0bNtS5rI4dO8JoNCI3Nxd9+vRpcCaZTAaj0WgxLSQkpEHLiomJwdGjRzFx4u3b0R49etT8d3d3dwQHB+PAgQPo16+fefqBAwfQrVu3Bq2zwWRqoOeM6kfmoeqjzlI2AboS2+ZoRHKTfReMmBOjW2A3PBbzGAaGDYRY5PzjYoQNlyyY2kRHR0Ov1+Pzzz/HqFGjcODAAXz55Zd1vq558+Z4+umnMXHiRCxduhQdO3bEzZs3sXPnTrRr1w4jRoyo1/ojIiKwbds2XLhwAT4+PvDw8Kix1XKn/Px88/jNLZ6enlAoFEhISMCzzz6LLl26oFevXli/fj1Onz6NyMhI87wvv/wyFixYgKioKHTo0AGrV69GcnIyfvjhh3rlFUSTHtWPYR8Cqb9Vl036Hoc7zNleC6aFdwuMjByJ4U2Hw0/lxzoOcQFUMH9r3749PvnkE3zwwQd4/fXX0bdvX7z//vsWWwH3snr1aixevBjz5s3D9evX4evrix49emDkyJH1Xv+zzz6L3bt3o0uXLigrK8OuXbvQv3//e84fFxdXY9q6deswfvx4PP3007h8+TLmz58PrVaLxx9/HPHx8Thy5Ih53tmzZ6O4uBjz5s1Dbm4uWrVqhc2bN6NZs2b1ziwYqbL6pM12jwHF14HT/wckrwPyG/kIN4EoTGxP2r1TkDoIw5sOx8jIkTS2QmzO5Y4ic1WDBw9GYGAgvvvuO9ZRGi4nBbi4o/px9RBgRz/I75TUrA/mGNjdAbS5V3P0De2LvqF90cGvA118kjBDWzBOqKKiAl9++SWGDBkCsViMdevWITExETt27GAd7cEEtK5+9J4LaEuAy7uqyyYtESjNYp3OTGG07e2SlRIlegT1QJ/QPugT0geB6kCbrp+Qe6EtGCdUWVmJUaNG4eTJk9BqtYiJicFbb72FRx55hHU04WSfqS6by7uBG8mArphZlONNOiNeLNxdP9VSNdr6tkUH/w7o5N8JnQM6Qyam2zQT+0MFQ5wPzwP5l4AbJ28/sk8DVWV1v7YRnA1piydljVdwTdyaoL1fe3Tw74D2fu3RzKsZRC5wxWbi+KhgiGswmYC8v4AbJ6q3cAouAUVXq28toK9o1FVdDIjBIyrrrlIg4kQI0YQg0iMSkZ6R1X/+/dC48JWpiWOjgiGkPA8ougIUZVaXTlFmdfGU3ACqyqsf+orqP3ljnYu76hOB4e63D1VWSpTwVnjDV+kLP6Vf9Z+q6j99lb4IUAUgwiMCcrFcyHdJiM1RwRBiDYOuZumYjIBY+vdDBqNYikKFBiqJCgqJgnZnEZdFBUMIIUQQ9KsVIYQQQVDBEEIIEQQVDCGEEEFQwRBCCBEEFQwhhBBBUMEQQggRBBUMIYQQQVDBEEIIEQQVDCGEEEFQwRBCCBEEFQwhhBBBUMEQQggRBBUMIYQQQVDBEEIIEQQVDCGEEEFQwRBCCBEEFQwhhBBBUMEQQggRBBUMIYQQQVDBEEIIEQQVDCGEEEFQwRBCCBEEFQwhhBBBUMEQQggRBBUMIYQQQVDBEEIIEQQVDCGEEEFQwRBCCBEEFQwhhBBBUMEQQggRBBUMIYQQQVDBEEIIEQQVDCGEEEFQwRBCCBEEFQwhhBBBUMEQQggRBBUMIYQQQVDBEEIIEQQVDCGEEEH8PxleQ2jZmR+wAAAAAElFTkSuQmCC\n"
          },
          "metadata": {}
        }
      ]
    }
  ]
}