experiment_notebook.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<a href=\"https://colab.research.google.com/github/walkerdustin/genAI-concept-search/blob/main/seek.ipynb\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open Notebook In Google Colab\"/></a>  "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "# import all packages we need\n",
    "import pandas as pd\n",
    "import numpy as np\n",
    "import subprocess\n",
    "import pickle\n",
    "import pythonbible as bible\n",
    "import time\n",
    "import os\n",
    "import matplotlib.pyplot as plt"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>id</th>\n",
       "      <th>b</th>\n",
       "      <th>c</th>\n",
       "      <th>v</th>\n",
       "      <th>t</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1001001</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>In the beginning God{After \"God,\" the Hebrew h...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1001002</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>Now the earth was formless and empty. Darkness...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1001003</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>God said, \"Let there be light,\" and there was ...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>1001004</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>4</td>\n",
       "      <td>God saw the light, and saw that it was good. G...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>1001005</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>5</td>\n",
       "      <td>God called the light Day, and the darkness he ...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31097</th>\n",
       "      <td>66022017</td>\n",
       "      <td>66</td>\n",
       "      <td>22</td>\n",
       "      <td>17</td>\n",
       "      <td>The Spirit and the bride say, \"Come!\" He who h...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31098</th>\n",
       "      <td>66022018</td>\n",
       "      <td>66</td>\n",
       "      <td>22</td>\n",
       "      <td>18</td>\n",
       "      <td>I testify to everyone who hears the words of t...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31099</th>\n",
       "      <td>66022019</td>\n",
       "      <td>66</td>\n",
       "      <td>22</td>\n",
       "      <td>19</td>\n",
       "      <td>If anyone takes away from the words of the boo...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31100</th>\n",
       "      <td>66022020</td>\n",
       "      <td>66</td>\n",
       "      <td>22</td>\n",
       "      <td>20</td>\n",
       "      <td>He who testifies these things says, \"Yes, I co...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31101</th>\n",
       "      <td>66022021</td>\n",
       "      <td>66</td>\n",
       "      <td>22</td>\n",
       "      <td>21</td>\n",
       "      <td>The grace of the Lord Jesus Christ be with all...</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>31102 rows × 5 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "             id   b   c   v                                                  t\n",
       "0       1001001   1   1   1  In the beginning God{After \"God,\" the Hebrew h...\n",
       "1       1001002   1   1   2  Now the earth was formless and empty. Darkness...\n",
       "2       1001003   1   1   3  God said, \"Let there be light,\" and there was ...\n",
       "3       1001004   1   1   4  God saw the light, and saw that it was good. G...\n",
       "4       1001005   1   1   5  God called the light Day, and the darkness he ...\n",
       "...         ...  ..  ..  ..                                                ...\n",
       "31097  66022017  66  22  17  The Spirit and the bride say, \"Come!\" He who h...\n",
       "31098  66022018  66  22  18  I testify to everyone who hears the words of t...\n",
       "31099  66022019  66  22  19  If anyone takes away from the words of the boo...\n",
       "31100  66022020  66  22  20  He who testifies these things says, \"Yes, I co...\n",
       "31101  66022021  66  22  21  The grace of the Lord Jesus Christ be with all...\n",
       "\n",
       "[31102 rows x 5 columns]"
      ]
     },
     "execution_count": 2,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# get the data\n",
    "verses = pd.read_csv(\"https://raw.githubusercontent.com/walkerdustin/local_LLM_experiment/main/data/t_web.csv\")\n",
    "# prompts should be a list of strings \"{citation}: {text}\"\n",
    "verses"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array(['The book of the generation of Jesus Christ{Christ (Greek) and Messiah (Hebrew) both mean \"Anointed One\"}, the son of David, the son of Abraham.',\n",
       "       'Abraham became the father of Isaac. Isaac became the father of Jacob. Jacob became the father of Judah and his brothers.',\n",
       "       'Judah became the father of Perez and Zerah by Tamar. Perez became the father of Hezron. Hezron became the father of Ram.',\n",
       "       ...,\n",
       "       'Jesus came to them and spoke to them, saying, \"All authority has been given to me in heaven and on earth.',\n",
       "       'Therefore go, and make disciples of all nations, baptizing them in the name of the Father and of the Son and of the Holy Spirit,',\n",
       "       'teaching them to observe all things that I commanded you. Behold, I am with you always, even to the end of the age.\" Amen.'],\n",
       "      dtype=object)"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# extract the bible verses from the date into a list of strings\n",
    "mathew_web_ids = bible.convert_reference_to_verse_ids(bible.NormalizedReference(book=bible.Book.MATTHEW, start_chapter=1, start_verse=1, end_chapter=28, end_verse=20, end_book=None),)\n",
    "references = [bible.convert_verse_ids_to_references([id])[0] for id in list(mathew_web_ids)]\n",
    "vers_references = [bible.format_single_reference(reference) for reference in references]\n",
    "# get all the verses coresponding to the  ids\n",
    "verses = verses.loc[verses[\"id\"].isin(mathew_web_ids), \"t\"].values\n",
    "verses"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "1071"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(verses)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## calc how expensive this should be for different providers:\n",
    "I test the following providers:\n",
    "- Google Cloud\n",
    "    - text bison (PaLM 2)\n",
    "    - Gemini Pro\n",
    "- OpenAI\n",
    "    - GPT 3.5"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Total Number of words in prompt: 119\n",
      "Total number of words in all prompts for the whole chapter: 149_800\n",
      "average chars per word: 5.53\n"
     ]
    }
   ],
   "source": [
    "sys_prompt = \"\"\"You are an expert in the understanding and interpretation of the English language. You will get a single verse of the bible as an input and respond only with a single number from 0 to 10, representing\n",
    "how much the given bible verse says about the concept of seeking discomfort. It is irrelevant if the verse suggests that seeking discomfort is good or bad.\n",
    "For every prompt you will answer with just one number (0 to 10). 0 represents the lowest correlation to the concept of seeking comfort or \n",
    "seeking discomfort and 10 represents the highest correlation to the concept of seeking comfort or seeking discomfort.\n",
    "Only answer with a single number. Do NOT give any explanation or context.\"\"\"\n",
    "print(f\"Total Number of words in prompt: {len(sys_prompt.split())}\")\n",
    "total_input_words = len(\"\".join(verses).split()) + len(sys_prompt.split())*len(verses)\n",
    "print(f\"Total number of words in all prompts for the whole chapter: {total_input_words:_}\")\n",
    "print(f\"average chars per word: { len(''.join(verses))/len(''.join(verses).split()):.2f}\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## now calculate the tokens with openAIs tokenizer tiktoken\n",
    "The actual cost depends on the tokens and not the words.  \n",
    "One token is about 0.8 words"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['gpt2', 'r50k_base', 'p50k_base', 'p50k_edit', 'cl100k_base']"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "import tiktoken\n",
    "tiktoken.list_encoding_names()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'You are an expert in the understanding and interpretation of the English language. You will get a single verse of the bible as an input and respond only with a single number from 0 to 10, representing\\nhow much the given bible verse says about the concept of seeking discomfort. It is irrelevant if the verse suggests that seeking discomfort is good or bad.\\nFor every prompt you will answer with just one number (0 to 10). 0 represents the lowest correlation to the concept of seeking comfort or \\nseeking discomfort and 10 represents the highest correlation to the concept of seeking comfort or seeking discomfort.\\nOnly answer with a single number. Do NOT give any explanation or context.'"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "sys_prompt"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'Y\\no\\nu\\n \\na\\nr\\ne\\n \\na\\nn\\n \\ne\\nx\\np\\ne\\nr\\nt\\n \\ni\\nn\\n \\nt\\nh\\ne\\n \\nu\\nn\\nd\\ne\\nr\\ns\\nt\\na\\nn\\nd\\ni\\nn\\ng\\n \\na\\nn\\nd\\n \\ni\\nn\\nt\\ne\\nr\\np\\nr\\ne\\nt\\na\\nt\\ni\\no\\nn\\n \\no\\nf\\n \\nt\\nh\\ne\\n \\nE\\nn\\ng\\nl\\ni\\ns\\nh\\n \\nl\\na\\nn\\ng\\nu\\na\\ng\\ne\\n.\\n \\nY\\no\\nu\\n \\nw\\ni\\nl\\nl\\n \\ng\\ne\\nt\\n \\na\\n \\ns\\ni\\nn\\ng\\nl\\ne\\n \\nv\\ne\\nr\\ns\\ne\\n \\no\\nf\\n \\nt\\nh\\ne\\n \\nb\\ni\\nb\\nl\\ne\\n \\na\\ns\\n \\na\\nn\\n \\ni\\nn\\np\\nu\\nt\\n \\na\\nn\\nd\\n \\nr\\ne\\ns\\np\\no\\nn\\nd\\n \\no\\nn\\nl\\ny\\n \\nw\\ni\\nt\\nh\\n \\na\\n \\ns\\ni\\nn\\ng\\nl\\ne\\n \\nn\\nu\\nm\\nb\\ne\\nr\\n \\nf\\nr\\no\\nm\\n \\n0\\n \\nt\\no\\n \\n1\\n0\\n,\\n \\nr\\ne\\np\\nr\\ne\\ns\\ne\\nn\\nt\\ni\\nn\\ng\\n\\n\\nh\\no\\nw\\n \\nm\\nu\\nc\\nh\\n \\nt\\nh\\ne\\n \\ng\\ni\\nv\\ne\\nn\\n \\nb\\ni\\nb\\nl\\ne\\n \\nv\\ne\\nr\\ns\\ne\\n \\ns\\na\\ny\\ns\\n \\na\\nb\\no\\nu\\nt\\n \\nt\\nh\\ne\\n \\nc\\no\\nn\\nc\\ne\\np\\nt\\n \\no\\nf\\n \\ns\\ne\\ne\\nk\\ni\\nn\\ng\\n \\nd\\ni\\ns\\nc\\no\\nm\\nf\\no\\nr\\nt\\n.\\n \\nI\\nt\\n \\ni\\ns\\n \\ni\\nr\\nr\\ne\\nl\\ne\\nv\\na\\nn\\nt\\n \\ni\\nf\\n \\nt\\nh\\ne\\n \\nv\\ne\\nr\\ns\\ne\\n \\ns\\nu\\ng\\ng\\ne\\ns\\nt\\ns\\n \\nt\\nh\\na\\nt\\n \\ns\\ne\\ne\\nk\\ni\\nn\\ng\\n \\nd\\ni\\ns\\nc\\no\\nm\\nf\\no\\nr\\nt\\n \\ni\\ns\\n \\ng\\no\\no\\nd\\n \\no\\nr\\n \\nb\\na\\nd\\n.\\n\\n\\nF\\no\\nr\\n \\ne\\nv\\ne\\nr\\ny\\n \\np\\nr\\no\\nm\\np\\nt\\n \\ny\\no\\nu\\n \\nw\\ni\\nl\\nl\\n \\na\\nn\\ns\\nw\\ne\\nr\\n \\nw\\ni\\nt\\nh\\n \\nj\\nu\\ns\\nt\\n \\no\\nn\\ne\\n \\nn\\nu\\nm\\nb\\ne\\nr\\n \\n(\\n0\\n \\nt\\no\\n \\n1\\n0\\n)\\n.\\n \\n0\\n \\nr\\ne\\np\\nr\\ne\\ns\\ne\\nn\\nt\\ns\\n \\nt\\nh\\ne\\n \\nl\\no\\nw\\ne\\ns\\nt\\n \\nc\\no\\nr\\nr\\ne\\nl\\na\\nt\\ni\\no\\nn\\n \\nt\\no\\n \\nt\\nh\\ne\\n \\nc\\no\\nn\\nc\\ne\\np\\nt\\n \\no\\nf\\n \\ns\\ne\\ne\\nk\\ni\\nn\\ng\\n \\nc\\no\\nm\\nf\\no\\nr\\nt\\n \\no\\nr\\n \\n\\n\\ns\\ne\\ne\\nk\\ni\\nn\\ng\\n \\nd\\ni\\ns\\nc\\no\\nm\\nf\\no\\nr\\nt\\n \\na\\nn\\nd\\n \\n1\\n0\\n \\nr\\ne\\np\\nr\\ne\\ns\\ne\\nn\\nt\\ns\\n \\nt\\nh\\ne\\n \\nh\\ni\\ng\\nh\\ne\\ns\\nt\\n \\nc\\no\\nr\\nr\\ne\\nl\\na\\nt\\ni\\no\\nn\\n \\nt\\no\\n \\nt\\nh\\ne\\n \\nc\\no\\nn\\nc\\ne\\np\\nt\\n \\no\\nf\\n \\ns\\ne\\ne\\nk\\ni\\nn\\ng\\n \\nc\\no\\nm\\nf\\no\\nr\\nt\\n \\no\\nr\\n \\ns\\ne\\ne\\nk\\ni\\nn\\ng\\n \\nd\\ni\\ns\\nc\\no\\nm\\nf\\no\\nr\\nt\\n.\\n\\n\\nO\\nn\\nl\\ny\\n \\na\\nn\\ns\\nw\\ne\\nr\\n \\nw\\ni\\nt\\nh\\n \\na\\n \\ns\\ni\\nn\\ng\\nl\\ne\\n \\nn\\nu\\nm\\nb\\ne\\nr\\n.\\n \\nD\\no\\n \\nN\\nO\\nT\\n \\ng\\ni\\nv\\ne\\n \\na\\nn\\ny\\n \\ne\\nx\\np\\nl\\na\\nn\\na\\nt\\ni\\no\\nn\\n \\no\\nr\\n \\nc\\no\\nn\\nt\\ne\\nx\\nt\\n.'"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "\"\\n\".join(sys_prompt)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "total tokens in whole bible chapter: 29874\n",
      "openai tokenizer average words per token: 0.7481756711521724\n",
      "openai tokenizer average chars per token: 4.135268126129745\n"
     ]
    }
   ],
   "source": [
    "enc = tiktoken.get_encoding(\"cl100k_base\") # this is the encoding for GPT 3 and 4\n",
    "# https://platform.openai.com/tokenizer\n",
    "# encoding and decoding should be inverses of one another\n",
    "assert enc.decode(enc.encode(\"hello world\")) == \"hello world\"\n",
    "\n",
    "number_of_Tokens_in_whole_chapter = len(enc.encode(\"\\n\".join(verses)))\n",
    "print(f\"total tokens in whole bible chapter: {number_of_Tokens_in_whole_chapter}\")\n",
    "\n",
    "total_words = len(\"\".join(verses).split())\n",
    "total_chars = len(\"\".join(verses))\n",
    "print(f\"openai tokenizer average words per token: {total_words/number_of_Tokens_in_whole_chapter}\")\n",
    "print(f\"openai tokenizer average chars per token: {total_chars/number_of_Tokens_in_whole_chapter}\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Total cost for GPT 3.5 == 0.179814\n",
      "Total cost for GPT 4 == 5.39442\n",
      "Total cost for Gemini Pro == 0.0449535\n",
      "Total cost for Mistral-7B-v0.2 == 0.02547384\n",
      "Total cost for Mixtral-8X7B-v0.1 == 0.1091736\n"
     ]
    }
   ],
   "source": [
    "total_input_tokens = number_of_Tokens_in_whole_chapter + len(verses) * len(enc.encode(sys_prompt))\n",
    "total_output_tokens = len(verses)*2 # I expect only a 2 digit number as an output\n",
    "# cost for openAi GPT-3.5\n",
    "cost_for_gpt35 = total_input_tokens * 1/1_000_000 + total_output_tokens * 2/1_000_000\n",
    "print(f\"Total cost for GPT 3.5 == {cost_for_gpt35}\")\n",
    "# cost for openAi GPT-4\n",
    "cost_for_gpt4 = total_input_tokens * 30/1_000_000 + total_output_tokens * 60/1_000_000\n",
    "print(f\"Total cost for GPT 4 == {cost_for_gpt4}\")\n",
    "# cost for google vertex-ai gemini pro and PaLM 2\n",
    "cost_for_gemini_pro = total_input_tokens * 0.25/1_000_000 + total_output_tokens * 0.5/1_000_000\n",
    "print(f\"Total cost for Gemini Pro == {cost_for_gemini_pro}\")\n",
    "# cost for mistral.ai Mistral-7B-v0.2\n",
    "cost_for_mistral7B = total_input_tokens * 0.14/1_000_000 + total_output_tokens * 0.42/1_000_000\n",
    "print(f\"Total cost for Mistral-7B-v0.2 == {cost_for_mistral7B}\")\n",
    "# cost for Mistral.ai Mixtral-8X7B-v0.1, a sparse mixture of experts model with 12B active parameters.\n",
    "cost_for_Mixtral8X7B = total_input_tokens * 0.6/1_000_000 + total_output_tokens * 1.8/1_000_000\n",
    "print(f\"Total cost for Mixtral-8X7B-v0.1 == {cost_for_Mixtral8X7B}\")\n",
    "# cost for mistral medium \n",
    "cost_for_MistralMedium = total_input_tokens * 2.5/1_000_000 + total_output_tokens * 7.5/1_000_000"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1000x600 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# import the necessary libraries\n",
    "import matplotlib.pyplot as plt\n",
    "\n",
    "# create a colorfull barplot for the costs\n",
    "costs = [cost_for_mistral7B, cost_for_Mixtral8X7B, cost_for_MistralMedium ,cost_for_gemini_pro, cost_for_gpt35, cost_for_gpt4, ]\n",
    "labels = [\"Mistral-7B\", \"Mixtral-8X7B\", \"Mistral-Medium\", \"Gemini Pro\", \"OpenAI GPT-3.5\", \"OpenAI GPT-4\"]\n",
    "colors = ['blue', 'green', 'red', 'orange', 'purple', 'gray']\n",
    "\n",
    "plt.figure(figsize=(10, 6))  # Adjust the figure size\n",
    "plt.bar(labels, costs, color=colors, width=0.5)  # Adjust bar width and add colors\n",
    "plt.ylabel(\"Costs in $\")\n",
    "plt.title(\"Cost for GenAI concept search in the bible chapter Matthew\")\n",
    "\n",
    "# Add data labels above the bars\n",
    "for i, cost in enumerate(costs):\n",
    "    plt.text(i, cost+ 0.1, f\"${cost:.2f}\", ha='center', fontsize=12)\n",
    "\n",
    "plt.ylim(0, max(costs) + 2)  # Adjust the y-axis limit for better visualization\n",
    "\n",
    "# Rotate x-axis labels for better readability\n",
    "plt.xticks(ha='center', fontsize=12)\n",
    "\n",
    "# Add a grid for better reference\n",
    "plt.grid(axis='y', linestyle='--', alpha=0.7)\n",
    "\n",
    "# Save the plot as \"costs.png\" with high DPI for better quality\n",
    "plt.savefig(\"images/costs.png\", dpi=300)\n",
    "plt.tight_layout()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## get responses from different models"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [],
   "source": [
    "# This is the Prompt Template I use for all Models\n",
    "system_prompt = \"\"\"You are an expert in the understanding and interpretation of the English language. You will get a single verse of the bible as an input and respond only with a single number from 0 to 10, representing\n",
    "how much the given bible verse says about the concept of seeking discomfort. It is irrelevant if the verse suggests that seeking discomfort is good or bad.\n",
    "For every prompt you will answer with just one number (0 to 10). 0 represents the lowest correlation to the concept of seeking comfort or \n",
    "seeking discomfort and 10 represents the highest correlation to the concept of seeking comfort or seeking discomfort.\n",
    "Only answer with a single number. Do NOT give any explanation or context.\n",
    "\n",
    "input: \"\"\""
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [],
   "source": [
    "# construct a dataframe for all the verses and respones\n",
    "experiment_df = pd.DataFrame({\n",
    "    \"vers_text\": verses,\n",
    "    \"bible_vers_id\":mathew_web_ids,\n",
    "     \"vers_references\":vers_references\n",
    "    })\n",
    "experiment_df\n",
    "times_df = pd.DataFrame(columns=[\"name\", \"sec\"])\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [],
   "source": [
    "# read in the experiment df from the previous run because there the answers are already appended\n",
    "# check if file is availabel\n",
    "if os.path.isfile('data/experiment_df.json'):\n",
    "    # read from csv\n",
    "    experiment_df = pd.read_json(\"data/experiment_df.json\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>vers_text</th>\n",
       "      <th>bible_vers_id</th>\n",
       "      <th>vers_references</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>The book of the generation of Jesus Christ{Chr...</td>\n",
       "      <td>40001001</td>\n",
       "      <td>Matthew 1:1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>Abraham became the father of Isaac. Isaac beca...</td>\n",
       "      <td>40001002</td>\n",
       "      <td>Matthew 1:2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>Judah became the father of Perez and Zerah by ...</td>\n",
       "      <td>40001003</td>\n",
       "      <td>Matthew 1:3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>Ram became the father of Amminadab. Amminadab ...</td>\n",
       "      <td>40001004</td>\n",
       "      <td>Matthew 1:4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>Salmon became the father of Boaz by Rahab. Boa...</td>\n",
       "      <td>40001005</td>\n",
       "      <td>Matthew 1:5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1066</th>\n",
       "      <td>But the eleven disciples went into Galilee, to...</td>\n",
       "      <td>40028016</td>\n",
       "      <td>Matthew 28:16</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1067</th>\n",
       "      <td>When they saw him, they bowed down to him, but...</td>\n",
       "      <td>40028017</td>\n",
       "      <td>Matthew 28:17</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1068</th>\n",
       "      <td>Jesus came to them and spoke to them, saying, ...</td>\n",
       "      <td>40028018</td>\n",
       "      <td>Matthew 28:18</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1069</th>\n",
       "      <td>Therefore go, and make disciples of all nation...</td>\n",
       "      <td>40028019</td>\n",
       "      <td>Matthew 28:19</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1070</th>\n",
       "      <td>teaching them to observe all things that I com...</td>\n",
       "      <td>40028020</td>\n",
       "      <td>Matthew 28:20</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>1071 rows × 3 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                                              vers_text  bible_vers_id  \\\n",
       "0     The book of the generation of Jesus Christ{Chr...       40001001   \n",
       "1     Abraham became the father of Isaac. Isaac beca...       40001002   \n",
       "2     Judah became the father of Perez and Zerah by ...       40001003   \n",
       "3     Ram became the father of Amminadab. Amminadab ...       40001004   \n",
       "4     Salmon became the father of Boaz by Rahab. Boa...       40001005   \n",
       "...                                                 ...            ...   \n",
       "1066  But the eleven disciples went into Galilee, to...       40028016   \n",
       "1067  When they saw him, they bowed down to him, but...       40028017   \n",
       "1068  Jesus came to them and spoke to them, saying, ...       40028018   \n",
       "1069  Therefore go, and make disciples of all nation...       40028019   \n",
       "1070  teaching them to observe all things that I com...       40028020   \n",
       "\n",
       "     vers_references  \n",
       "0        Matthew 1:1  \n",
       "1        Matthew 1:2  \n",
       "2        Matthew 1:3  \n",
       "3        Matthew 1:4  \n",
       "4        Matthew 1:5  \n",
       "...              ...  \n",
       "1066   Matthew 28:16  \n",
       "1067   Matthew 28:17  \n",
       "1068   Matthew 28:18  \n",
       "1069   Matthew 28:19  \n",
       "1070   Matthew 28:20  \n",
       "\n",
       "[1071 rows x 3 columns]"
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "experiment_df"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### get responses from PaLM 2 (text-bison)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [],
   "source": [
    "import vertexai\n",
    "from vertexai.language_models import TextGenerationModel\n",
    "# this authentication works in online google colab notebook\n",
    "# from google.colab import auth\n",
    "# auth.authenticate_user()\n",
    "\n",
    "# !gcloud auth login  ## authentication should work in jupyter notebook\n",
    "\n",
    "# this authentication works in a local ipython notebook and local python\n",
    "os.environ['GOOGLE_APPLICATION_CREDENTIALS'] = '/mnt/c/Git_Repos/try-out-vertex-ai-20aaf5516b7c.json'\n",
    "vertexai.init(project=\"try-out-vertex-ai\", location=\"us-central1\", service_account=\"for-local-vertex-ai@try-out-vertex-ai.iam.gserviceaccount.com\")\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Example Prompt with PaLM 2"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Response from Model:  0\n"
     ]
    }
   ],
   "source": [
    "parameters = {\n",
    "    \"candidate_count\": 1,\n",
    "    \"max_output_tokens\": 2,\n",
    "    \"temperature\": 0,\n",
    "    \"top_k\": 1\n",
    "}\n",
    "model = TextGenerationModel.from_pretrained(\"text-bison\")\n",
    "response = model.predict(system_prompt + experiment_df[\"vers_text\"].iloc[641], **parameters)\n",
    "print(f\"Response from Model: {response.text}\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [],
   "source": [
    "if response.errors:\n",
    "    print(f\"Model had some errors: {response.errors}\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[ 0]\n",
      "()\n",
      "GroundingMetadata(citations=[], search_queries=[])\n",
      "False\n",
      "Prediction(predictions=[{'content': ' 0', 'citationMetadata': {'citations': []}, 'safetyAttributes': {'blocked': False, 'safetyRatings': [{'probabilityScore': 0.2, 'severityScore': 0.3, 'severity': 'LOW', 'category': 'Dangerous Content'}, {'probabilityScore': 0.2, 'severityScore': 0.1, 'severity': 'NEGLIGIBLE', 'category': 'Harassment'}, {'probabilityScore': 0.2, 'severityScore': 0.2, 'severity': 'NEGLIGIBLE', 'category': 'Hate Speech'}, {'probabilityScore': 0.1, 'severityScore': 0.1, 'severity': 'NEGLIGIBLE', 'category': 'Sexually Explicit'}], 'scores': [0.2, 0.3, 0.6, 0.1, 0.2, 0.1, 0.2, 0.1, 1.0, 0.1, 0.1, 0.1], 'categories': ['Derogatory', 'Finance', 'Health', 'Illicit Drugs', 'Insult', 'Legal', 'Politics', 'Public Safety', 'Religion & Belief', 'Sexual', 'Toxic', 'War & Conflict']}}], deployed_model_id='', model_version_id='', model_resource_name='', explanations=None)\n",
      "{'Derogatory': 0.2, 'Finance': 0.3, 'Health': 0.6, 'Illicit Drugs': 0.1, 'Insult': 0.2, 'Legal': 0.1, 'Politics': 0.2, 'Public Safety': 0.1, 'Religion & Belief': 1.0, 'Sexual': 0.1, 'Toxic': 0.1, 'War & Conflict': 0.1}\n"
     ]
    }
   ],
   "source": [
    "# 'candidates',\n",
    "# 'errors',\n",
    "# 'grounding_metadata',\n",
    "# 'is_blocked',\n",
    "# 'raw_prediction_response',\n",
    "# 'safety_attributes',\n",
    "# 'text'\n",
    "print(response.candidates)\n",
    "print(response.errors)\n",
    "print(response.grounding_metadata)\n",
    "print(response.is_blocked)\n",
    "print(response.raw_prediction_response)\n",
    "print(response.safety_attributes)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  1\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "# the current counter is 60\n",
      "waiting\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  8\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  1\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  5\n",
      "Response from Model:  8\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  5\n",
      "# the current counter is 120\n",
      "waiting\n",
      "Response from Model:  5\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  1\n",
      "Response from Model:  2\n",
      "Response from Model:  5\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  5\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  2\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  5\n",
      "# the current counter is 180\n",
      "waiting\n",
      "Response from Model:  7\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  5\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  1\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  2\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "# the current counter is 240\n",
      "waiting\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  1\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  2\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  2\n",
      "Response from Model:  1\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "# the current counter is 300\n",
      "waiting\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  3\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  5\n",
      "Response from Model:  5\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  7\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  1\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "# the current counter is 360\n",
      "waiting\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  2\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  5\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  2\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "# the current counter is 420\n",
      "waiting\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "# the current counter is 480\n",
      "waiting\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  1\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "# the current counter is 540\n",
      "waiting\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  2\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  8\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  7\n",
      "Response from Model:  5\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  5\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "# the current counter is 600\n",
      "waiting\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  3\n",
      "Response from Model:  0\n",
      "Response from Model:  2\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "# the current counter is 660\n",
      "waiting\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  3\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "# the current counter is 720\n",
      "waiting\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "# the current counter is 780\n",
      "waiting\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  5\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  1\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "# the current counter is 840\n",
      "waiting\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "# the current counter is 900\n",
      "waiting\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  1\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  2\n",
      "Response from Model:  0\n",
      "Response from Model:  2\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "# the current counter is 960\n",
      "waiting\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "# the current counter is 1020\n",
      "waiting\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "Response from Model:  0\n",
      "all 1071 prompts took: 549.71 seconds\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/tmp/ipykernel_21183/3734296186.py:43: FutureWarning: The behavior of DataFrame concatenation with empty or all-NA entries is deprecated. In a future version, this will no longer exclude empty or all-NA columns when determining the result dtypes. To retain the old behavior, exclude the relevant entries before the concat operation.\n",
      "  times_df = pd.concat([times_df, pd.DataFrame([{'name': 'text_bison', 'sec': total_time_for_palm2}])])\n"
     ]
    }
   ],
   "source": [
    "# if the column text_bison_ratings already exists, then read skip this execution\n",
    "if \"text_bison_ratings\" in list(experiment_df.columns):\n",
    "    print(\"text_bison_ratings already exists\")\n",
    "else:\n",
    "    # get the score for each prompt\n",
    "    WAIT_TIME_AGAINST_RATE_LIMIT_IN_SECONDS = 1\n",
    "    text_bison_responses = []\n",
    "    text_bison_response_objects = []\n",
    "    start_timestamp = time.time()\n",
    "    last_start_timestamp = time.time()\n",
    "    counter = 0\n",
    "    for prompt in verses:\n",
    "        counter += 1\n",
    "        if counter%60 == 0:\n",
    "            print(\"# the current counter is\", counter)\n",
    "            print(\"waiting\")\n",
    "            while time.time() - last_start_timestamp < WAIT_TIME_AGAINST_RATE_LIMIT_IN_SECONDS:\n",
    "                time.sleep(0.001)\n",
    "            last_start_timestamp = time.time()\n",
    "        response = model.predict(system_prompt + prompt, **parameters)\n",
    "        print(f\"Response from Model: {response.text}\")\n",
    "        result = None\n",
    "        if response.errors:\n",
    "            result = -1\n",
    "        if response.is_blocked:\n",
    "            result = -2\n",
    "        if not response.errors and not response.is_blocked:\n",
    "            result = response.text\n",
    "        text_bison_responses.append(response.text)\n",
    "        text_bison_response_objects.append(response)\n",
    "\n",
    "    total_time_for_palm2 = time.time() - start_timestamp\n",
    "    print(f\"all {len(verses)} prompts took: {total_time_for_palm2:.2f} seconds\")\n",
    "\n",
    "    # strip all responses\n",
    "    text_bison_responses_int = [int(response.strip()) if response.strip() != \"\" else -3 for response in text_bison_responses ]\n",
    "    \n",
    "    # add text_bison_responses_int as colmn to df\n",
    "    experiment_df.insert(2,\"text_bison_ratings\", text_bison_responses_int)\n",
    "    experiment_df.insert(2,\"text_bison_response_objects\", text_bison_response_objects)\n",
    "    \n",
    "    # save time in times_df\n",
    "    times_df = pd.concat([times_df, pd.DataFrame([{'name': 'text_bison', 'sec': total_time_for_palm2}])])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### get responses from gemini pro"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Greetings,\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "candidates {\n",
       "  content {\n",
       "    role: \"model\"\n",
       "    parts {\n",
       "      text: \"Greetings,\"\n",
       "    }\n",
       "  }\n",
       "  finish_reason: MAX_TOKENS\n",
       "  safety_ratings {\n",
       "    category: HARM_CATEGORY_HATE_SPEECH\n",
       "    probability: NEGLIGIBLE\n",
       "  }\n",
       "  safety_ratings {\n",
       "    category: HARM_CATEGORY_DANGEROUS_CONTENT\n",
       "    probability: NEGLIGIBLE\n",
       "  }\n",
       "  safety_ratings {\n",
       "    category: HARM_CATEGORY_HARASSMENT\n",
       "    probability: NEGLIGIBLE\n",
       "  }\n",
       "  safety_ratings {\n",
       "    category: HARM_CATEGORY_SEXUALLY_EXPLICIT\n",
       "    probability: NEGLIGIBLE\n",
       "  }\n",
       "}\n",
       "usage_metadata {\n",
       "  prompt_token_count: 3\n",
       "  candidates_token_count: 2\n",
       "  total_token_count: 5\n",
       "}"
      ]
     },
     "execution_count": 21,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from vertexai.preview.generative_models import GenerativeModel\n",
    "model = GenerativeModel(\"gemini-pro\")\n",
    "generation_config={\n",
    "  \"candidate_count\": 1,\n",
    "  \"max_output_tokens\": 2,\n",
    "  \"temperature\": 0,\n",
    "  \"top_k\": 1\n",
    "}\n",
    "prompt = \"Hello Gemini!\"\n",
    "responses = model.generate_content(prompt, generation_config=generation_config)\n",
    "print(responses.candidates[0].text)\n",
    "responses"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "# the current counter is 60\n",
      "waiting\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "8\n",
      "2\n",
      "2\n",
      "3\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "2\n",
      "0\n",
      "7\n",
      "0\n",
      "0\n",
      "0\n",
      "8\n",
      "10\n",
      "0\n",
      "3\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "-1\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "-1\n",
      "7\n",
      "# the current counter is 120\n",
      "waiting\n",
      "8\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "2\n",
      "5\n",
      "5\n",
      "0\n",
      "-1\n",
      "7\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "-1\n",
      "0\n",
      "5\n",
      "# the current counter is 180\n",
      "waiting\n",
      "7\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "7\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "2\n",
      "0\n",
      "5\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "# the current counter is 240\n",
      "waiting\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "7\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "-1\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "5\n",
      "2\n",
      "0\n",
      "0\n",
      "0\n",
      "# the current counter is 300\n",
      "waiting\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "7\n",
      "0\n",
      "0\n",
      "0\n",
      "7\n",
      "8\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "7\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "2\n",
      "2\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "# the current counter is 360\n",
      "waiting\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "5\n",
      "0\n",
      "0\n",
      "0\n",
      "# the current counter is 420\n",
      "waiting\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "# the current counter is 480\n",
      "waiting\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "-1\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "# the current counter is 540\n",
      "waiting\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "7\n",
      "0\n",
      "0\n",
      "8\n",
      "7\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "8\n",
      "8\n",
      "0\n",
      "0\n",
      "2\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "# the current counter is 600\n",
      "waiting\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "-1\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "5\n",
      "0\n",
      "3\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "# the current counter is 660\n",
      "waiting\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "5\n",
      "0\n",
      "0\n",
      "3\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "# the current counter is 720\n",
      "waiting\n",
      "-1\n",
      "0\n",
      "0\n",
      "-1\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "-1\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "-1\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "# the current counter is 780\n",
      "waiting\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "-1\n",
      "0\n",
      "-1\n",
      "0\n",
      "-1\n",
      "0\n",
      "-1\n",
      "0\n",
      "0\n",
      "0\n",
      "-1\n",
      "0\n",
      "-1\n",
      "0\n",
      "-1\n",
      "-1\n",
      "-1\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "7\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "2\n",
      "0\n",
      "0\n",
      "-1\n",
      "-1\n",
      "0\n",
      "0\n",
      "5\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "# the current counter is 840\n",
      "waiting\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "-1\n",
      "0\n",
      "-1\n",
      "0\n",
      "-1\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "-1\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "# the current counter is 900\n",
      "waiting\n",
      "-1\n",
      "0\n",
      "-1\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "-1\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "2\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "# the current counter is 960\n",
      "waiting\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "-1\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "# the current counter is 1020\n",
      "waiting\n",
      "0\n",
      "0\n",
      "-1\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "all 1071 prompts took: 456.11 seconds\n"
     ]
    }
   ],
   "source": [
    "# if the column gemini_pro_ratings already exists, then read skip this execution\n",
    "if \"gemini_pro_ratings\" in list(experiment_df.columns):\n",
    "    print(\"gemini_pro_ratings already exists\")\n",
    "else:\n",
    "    # get the score for each prompt\n",
    "    WAIT_TIME_AGAINST_RATE_LIMIT_IN_SECONDS = 1\n",
    "    gemini_pro_responses = []\n",
    "    gemini_pro_response_objects = []\n",
    "    start_timestamp = time.time()\n",
    "    last_start_timestamp = time.time()\n",
    "    counter = 0\n",
    "    for prompt in verses:\n",
    "        counter += 1\n",
    "        if counter%60 == 0:\n",
    "            print(\"# the current counter is\", counter)\n",
    "            print(\"waiting\")\n",
    "            while time.time() - last_start_timestamp < WAIT_TIME_AGAINST_RATE_LIMIT_IN_SECONDS:\n",
    "                time.sleep(0.001)\n",
    "            last_start_timestamp = time.time()\n",
    "        try:\n",
    "            response = model.generate_content(system_prompt + prompt, generation_config=generation_config)\n",
    "        except Exception as e:\n",
    "            text = \"-3\"\n",
    "        else:\n",
    "            try:\n",
    "                text = response.candidates[0].text\n",
    "            except ValueError:\n",
    "                text = \"-1\"\n",
    "        print(text)\n",
    "        gemini_pro_responses.append(text)\n",
    "        gemini_pro_response_objects.append(response)\n",
    "    \n",
    "    total_time_for_gemini_pro = time.time() - start_timestamp\n",
    "    print(f\"all {len(verses)} prompts took: {total_time_for_gemini_pro:.2f} seconds\")\n",
    "\n",
    "    gemini_pro_responses_int = [int(response.strip()) if response.strip() != \"\" else -1 for response in gemini_pro_responses ]\n",
    "\n",
    "    experiment_df.insert(2,\"gemini_pro_ratings\", gemini_pro_responses_int)\n",
    "    experiment_df.insert(2,\"gemini_pro_response_objects\", gemini_pro_response_objects)\n",
    "    \n",
    "    # save time in times_df\n",
    "    times_df = pd.concat([times_df, pd.DataFrame([{'name': 'gemini_pro', 'sec': total_time_for_gemini_pro}])])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>vers_text</th>\n",
       "      <th>gemini_pro_ratings</th>\n",
       "      <th>text_bison_ratings</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>100</th>\n",
       "      <td>\"Blessed are you when people reproach you, persecute you, and say all kinds of evil against you falsely, for my sake.</td>\n",
       "      <td>10</td>\n",
       "      <td>8</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>590</th>\n",
       "      <td>If your hand or your foot causes you to stumble, cut it off, and cast it from you. It is better for you to enter into life maimed or crippled, rather than having two hands or two feet to be cast into the eternal fire.</td>\n",
       "      <td>8</td>\n",
       "      <td>7</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>591</th>\n",
       "      <td>If your eye causes you to stumble, pluck it out, and cast it from you. It is better for you to enter into life with one eye, rather than having two eyes to be cast into the Gehenna{or, Hell} of fire.</td>\n",
       "      <td>8</td>\n",
       "      <td>5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>65</th>\n",
       "      <td>Then Jesus was led up by the Spirit into the wilderness to be tempted by the devil.</td>\n",
       "      <td>8</td>\n",
       "      <td>8</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>311</th>\n",
       "      <td>He who finds his life will lose it; and he who loses his life for my sake will find it.</td>\n",
       "      <td>8</td>\n",
       "      <td>5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>99</th>\n",
       "      <td>Blessed are those who have been persecuted for righteousness' sake, For theirs is the Kingdom of Heaven.</td>\n",
       "      <td>8</td>\n",
       "      <td>5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>551</th>\n",
       "      <td>Then Jesus said to his disciples, \"If anyone desires to come after me, let him deny himself, and take up his cross, and follow me.</td>\n",
       "      <td>8</td>\n",
       "      <td>8</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>119</th>\n",
       "      <td>If your right hand causes you to stumble, cut it off, and throw it away from you: for it is profitable for you that one of your members should perish, and not your whole body be thrown into Gehenna.</td>\n",
       "      <td>8</td>\n",
       "      <td>5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>179</th>\n",
       "      <td>For everyone who asks receives. He who seeks finds. To him who knocks it will be opened.</td>\n",
       "      <td>7</td>\n",
       "      <td>7</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>310</th>\n",
       "      <td>He who doesn't take his cross and follow after me, isn't worthy of me.</td>\n",
       "      <td>7</td>\n",
       "      <td>5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>306</th>\n",
       "      <td>\"Don't think that I came to send peace on the earth. I didn't come to send peace, but a sword.</td>\n",
       "      <td>7</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>95</th>\n",
       "      <td>Blessed are those who hunger and thirst after righteousness, For they shall be filled.</td>\n",
       "      <td>7</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>326</th>\n",
       "      <td>From the days of John the Baptizer until now, the Kingdom of Heaven suffers violence, and the violent take it by force.</td>\n",
       "      <td>7</td>\n",
       "      <td>7</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>184</th>\n",
       "      <td>\"Enter in by the narrow gate; for wide is the gate and broad is the way that leads to destruction, and many are those who enter in by it.</td>\n",
       "      <td>7</td>\n",
       "      <td>5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>118</th>\n",
       "      <td>If your right eye causes you to stumble, pluck it out and throw it away from you. For it is more profitable for you that one of your members should perish, than for your whole body to be cast into Gehenna.</td>\n",
       "      <td>7</td>\n",
       "      <td>5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>282</th>\n",
       "      <td>Take no bag for your journey, neither two coats, nor shoes, nor staff: for the laborer is worthy of his food.</td>\n",
       "      <td>7</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>552</th>\n",
       "      <td>For whoever desires to save his life will lose it, and whoever will lose his life for my sake will find it.</td>\n",
       "      <td>7</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>133</th>\n",
       "      <td>But I tell you, love your enemies, bless those who curse you, do good to those who hate you, and pray for those who mistreat you and persecute you,</td>\n",
       "      <td>7</td>\n",
       "      <td>5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>548</th>\n",
       "      <td>From that time, Jesus began to show his disciples that he must go to Jerusalem and suffer many things from the elders, chief priests, and scribes, and be killed, and the third day be raised up.</td>\n",
       "      <td>7</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>807</th>\n",
       "      <td>Therefore, behold, I send to you prophets, wise men, and scribes. Some of them you will kill and crucify; and some of them you will scourge in your synagogues, and persecute from city to city;</td>\n",
       "      <td>7</td>\n",
       "      <td>5</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                                                                                                                                                                                                     vers_text  \\\n",
       "100  \"Blessed are you when people reproach you, persecute you, and say all kinds of evil against you falsely, for my sake.                                                                                                       \n",
       "590  If your hand or your foot causes you to stumble, cut it off, and cast it from you. It is better for you to enter into life maimed or crippled, rather than having two hands or two feet to be cast into the eternal fire.   \n",
       "591  If your eye causes you to stumble, pluck it out, and cast it from you. It is better for you to enter into life with one eye, rather than having two eyes to be cast into the Gehenna{or, Hell} of fire.                     \n",
       "65   Then Jesus was led up by the Spirit into the wilderness to be tempted by the devil.                                                                                                                                         \n",
       "311  He who finds his life will lose it; and he who loses his life for my sake will find it.                                                                                                                                     \n",
       "99   Blessed are those who have been persecuted for righteousness' sake, For theirs is the Kingdom of Heaven.                                                                                                                    \n",
       "551  Then Jesus said to his disciples, \"If anyone desires to come after me, let him deny himself, and take up his cross, and follow me.                                                                                          \n",
       "119  If your right hand causes you to stumble, cut it off, and throw it away from you: for it is profitable for you that one of your members should perish, and not your whole body be thrown into Gehenna.                      \n",
       "179  For everyone who asks receives. He who seeks finds. To him who knocks it will be opened.                                                                                                                                    \n",
       "310  He who doesn't take his cross and follow after me, isn't worthy of me.                                                                                                                                                      \n",
       "306  \"Don't think that I came to send peace on the earth. I didn't come to send peace, but a sword.                                                                                                                              \n",
       "95   Blessed are those who hunger and thirst after righteousness, For they shall be filled.                                                                                                                                      \n",
       "326  From the days of John the Baptizer until now, the Kingdom of Heaven suffers violence, and the violent take it by force.                                                                                                     \n",
       "184  \"Enter in by the narrow gate; for wide is the gate and broad is the way that leads to destruction, and many are those who enter in by it.                                                                                   \n",
       "118  If your right eye causes you to stumble, pluck it out and throw it away from you. For it is more profitable for you that one of your members should perish, than for your whole body to be cast into Gehenna.               \n",
       "282  Take no bag for your journey, neither two coats, nor shoes, nor staff: for the laborer is worthy of his food.                                                                                                               \n",
       "552  For whoever desires to save his life will lose it, and whoever will lose his life for my sake will find it.                                                                                                                 \n",
       "133  But I tell you, love your enemies, bless those who curse you, do good to those who hate you, and pray for those who mistreat you and persecute you,                                                                         \n",
       "548  From that time, Jesus began to show his disciples that he must go to Jerusalem and suffer many things from the elders, chief priests, and scribes, and be killed, and the third day be raised up.                           \n",
       "807  Therefore, behold, I send to you prophets, wise men, and scribes. Some of them you will kill and crucify; and some of them you will scourge in your synagogues, and persecute from city to city;                            \n",
       "\n",
       "     gemini_pro_ratings  text_bison_ratings  \n",
       "100  10                  8                   \n",
       "590  8                   7                   \n",
       "591  8                   5                   \n",
       "65   8                   8                   \n",
       "311  8                   5                   \n",
       "99   8                   5                   \n",
       "551  8                   8                   \n",
       "119  8                   5                   \n",
       "179  7                   7                   \n",
       "310  7                   5                   \n",
       "306  7                   3                   \n",
       "95   7                   0                   \n",
       "326  7                   7                   \n",
       "184  7                   5                   \n",
       "118  7                   5                   \n",
       "282  7                   2                   \n",
       "552  7                   0                   \n",
       "133  7                   5                   \n",
       "548  7                   2                   \n",
       "807  7                   5                   "
      ]
     },
     "execution_count": 23,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pd.set_option('display.max_colwidth', 0)\n",
    "experiment_df.sort_values(\"gemini_pro_ratings\", ascending=False)[[\"vers_text\", \"gemini_pro_ratings\", \"text_bison_ratings\"]].head(20)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## use Openai Api"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [],
   "source": [
    "# set the openai Api key in the enviroment with\n",
    "# setx OPENAI_API_KEY \"your-api-key-here\"\n",
    "\n",
    "# or in python with\n",
    "# os.environ[\"OPENAI_API_KEY\"] = \"your-api-key-here\""
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [],
   "source": [
    "system_prompt = \"\"\"You are an expert in the understanding and interpretation of the English language. You will get a single verse of the bible as an input and respond only with a single number from 0 to 10, representing\n",
    "how much the given bible verse says about the concept of seeking discomfort. It is irrelevant if the verse suggests that seeking discomfort is good or bad.\n",
    "For every prompt you will answer with just one number (0 to 10). 0 represents the lowest correlation to the concept of seeking comfort or \n",
    "seeking discomfort and 10 represents the highest correlation to the concept of seeking comfort or seeking discomfort.\n",
    "Only answer with a single number. Do NOT give any explanation or context.\n",
    "\"\"\""
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "ChatCompletionMessage(content='2', role='assistant', function_call=None, tool_calls=None)\n"
     ]
    }
   ],
   "source": [
    "from openai import OpenAI\n",
    "client = OpenAI()\n",
    "# defaults to getting the key using os.environ.get(\"OPENAI_API_KEY\")\n",
    "# if you saved the key under a different environment variable name, you can do something like:\n",
    "# client = OpenAI(\n",
    "#   api_key=os.environ.get(\"CUSTOM_ENV_NAME\"),\n",
    "# )\n",
    "def get_GPT3_5_response(prompt):\n",
    "  completion = client.chat.completions.create(\n",
    "    model=\"gpt-3.5-turbo\",\n",
    "    max_tokens=2,\n",
    "    temperature=0,\n",
    "    messages=[\n",
    "      {\"role\": \"system\", \"content\": system_prompt},\n",
    "      {\"role\": \"user\", \"content\": prompt}\n",
    "    ]\n",
    "  )\n",
    "  return completion\n",
    "print(get_GPT3_5_response(verses[30]).choices[0].message)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "8\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "ChatCompletion(id='chatcmpl-8v3IwbIA8m4hv5ob2YBRVv36DDrxk', choices=[Choice(finish_reason='stop', index=0, logprobs=None, message=ChatCompletionMessage(content='8', role='assistant', function_call=None, tool_calls=None))], created=1708608090, model='gpt-3.5-turbo-0125', object='chat.completion', system_fingerprint='fp_cbdb91ce3f', usage=CompletionUsage(completion_tokens=1, prompt_tokens=165, total_tokens=166))"
      ]
     },
     "execution_count": 27,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "completion = client.chat.completions.create(\n",
    "  model=\"gpt-3.5-turbo\",\n",
    "  messages=[\n",
    "    {\"role\": \"system\", \"content\": system_prompt},\n",
    "    {\"role\": \"user\", \"content\": verses[310]}\n",
    "  ]\n",
    ")\n",
    "print(completion.choices[0].message.content)\n",
    "completion"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [],
   "source": [
    "def convert_to_int(s):\n",
    "    try:\n",
    "        return int(s)\n",
    "    except ValueError:\n",
    "        return -1"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "2\n",
      "2\n",
      "2\n",
      "1\n",
      "0\n",
      "0\n",
      "2\n",
      "0\n",
      "2\n",
      "2\n",
      "2\n",
      "2\n",
      "0\n",
      "2\n",
      "2\n",
      "2\n",
      "2\n",
      "2\n",
      "2\n",
      "2\n",
      "8\n",
      "8\n",
      "1\n",
      "2\n",
      "0\n",
      "2\n",
      "0\n",
      "2\n",
      "1\n",
      "2\n",
      "1\n",
      "2\n",
      "2\n",
      "2\n",
      "8\n",
      "1\n",
      "1\n",
      "8\n",
      "2\n",
      "2\n",
      "2\n",
      "2\n",
      "# the current counter is 60\n",
      "2\n",
      "2\n",
      "2\n",
      "2\n",
      "2\n",
      "1\n",
      "10\n",
      "8\n",
      "2\n",
      "2\n",
      "8\n",
      "8\n",
      "2\n",
      "2\n",
      "2\n",
      "2\n",
      "2\n",
      "2\n",
      "2\n",
      "0\n",
      "2\n",
      "2\n",
      "2\n",
      "2\n",
      "2\n",
      "8\n",
      "2\n",
      "8\n",
      "2\n",
      "5\n",
      "1\n",
      "2\n",
      "0\n",
      "2\n",
      "2\n",
      "2\n",
      "7\n",
      "2\n",
      "2\n",
      "2\n",
      "8\n",
      "8\n",
      "2\n",
      "2\n",
      "2\n",
      "2\n",
      "2\n",
      "2\n",
      "2\n",
      "2\n",
      "2\n",
      "0\n",
      "2\n",
      "2\n",
      "2\n",
      "2\n",
      "2\n",
      "0\n",
      "1\n",
      "8\n",
      "# the current counter is 120\n",
      "8\n",
      "0\n",
      "1\n",
      "2\n",
      "1\n",
      "0\n",
      "1\n",
      "2\n",
      "0\n",
      "8\n",
      "8\n",
      "7\n",
      "2\n",
      "1\n",
      "2\n",
      "2\n",
      "8\n",
      "2\n",
      "2\n",
      "2\n",
      "2\n",
      "2\n",
      "2\n",
      "2\n",
      "2\n",
      "2\n",
      "1\n",
      "1\n",
      "2\n",
      "2\n",
      "2\n",
      "2\n",
      "2\n",
      "2\n",
      "2\n",
      "2\n",
      "2\n",
      "2\n",
      "2\n",
      "2\n",
      "2\n",
      "2\n",
      "2\n",
      "7\n",
      "2\n",
      "2\n",
      "2\n",
      "2\n",
      "2\n",
      "2\n",
      "2\n",
      "2\n",
      "2\n",
      "2\n",
      "2\n",
      "7\n",
      "7\n",
      "2\n",
      "2\n",
      "8\n",
      "# the current counter is 180\n",
      "8\n",
      "2\n",
      "2\n",
      "2\n",
      "2\n",
      "8\n",
      "8\n",
      "2\n",
      "7\n",
      "2\n",
      "2\n",
      "8\n",
      "1\n",
      "2\n",
      "2\n",
      "0\n",
      "2\n",
      "2\n",
      "2\n",
      "8\n",
      "1\n",
      "2\n",
      "1\n",
      "2\n",
      "2\n",
      "2\n",
      "1\n",
      "1\n",
      "1\n",
      "2\n",
      "2\n",
      "2\n",
      "2\n",
      "2\n",
      "2\n",
      "1\n",
      "2\n",
      "2\n",
      "0\n",
      "2\n",
      "2\n",
      "8\n",
      "2\n",
      "8\n",
      "1\n",
      "8\n",
      "2\n",
      "2\n",
      "2\n",
      "2\n",
      "2\n",
      "1\n",
      "2\n",
      "8\n",
      "2\n",
      "7\n",
      "1\n",
      "2\n",
      "1\n",
      "2\n",
      "# the current counter is 240\n",
      "5\n",
      "8\n",
      "1\n",
      "1\n",
      "2\n",
      "2\n",
      "2\n",
      "2\n",
      "2\n",
      "2\n",
      "2\n",
      "2\n",
      "2\n",
      "2\n",
      "2\n",
      "2\n",
      "7\n",
      "2\n",
      "2\n",
      "2\n",
      "2\n",
      "1\n",
      "2\n",
      "2\n",
      "2\n",
      "2\n",
      "1\n",
      "1\n",
      "2\n",
      "1\n",
      "2\n",
      "2\n",
      "2\n",
      "2\n",
      "2\n",
      "0\n",
      "0\n",
      "0\n",
      "2\n",
      "1\n",
      "2\n",
      "2\n",
      "8\n",
      "8\n",
      "2\n",
      "1\n",
      "2\n",
      "7\n",
      "2\n",
      "8\n",
      "8\n",
      "2\n",
      "2\n",
      "2\n",
      "2\n",
      "8\n",
      "8\n",
      "2\n",
      "2\n",
      "2\n",
      "# the current counter is 300\n",
      "2\n",
      "2\n",
      "2\n",
      "1\n",
      "2\n",
      "2\n",
      "2\n",
      "8\n",
      "7\n",
      "2\n",
      "8\n",
      "8\n",
      "8\n",
      "2\n",
      "2\n",
      "2\n",
      "1\n",
      "2\n",
      "2\n",
      "2\n",
      "2\n",
      "2\n",
      "2\n",
      "2\n",
      "2\n",
      "1\n",
      "2\n",
      "8\n",
      "2\n",
      "2\n",
      "2\n",
      "2\n",
      "3\n",
      "2\n",
      "3\n",
      "2\n",
      "2\n",
      "1\n",
      "2\n",
      "1\n",
      "2\n",
      "1\n",
      "0\n",
      "2\n",
      "2\n",
      "2\n",
      "2\n",
      "1\n",
      "2\n",
      "8\n",
      "2\n",
      "1\n",
      "2\n",
      "1\n",
      "1\n",
      "2\n",
      "2\n",
      "1\n",
      "2\n",
      "2\n",
      "# the current counter is 360\n",
      "2\n",
      "0\n",
      "0\n",
      "1\n",
      "1\n",
      "2\n",
      "1\n",
      "2\n",
      "1\n",
      "1\n",
      "2\n",
      "2\n",
      "2\n",
      "1\n",
      "8\n",
      "2\n",
      "0\n",
      "2\n",
      "2\n",
      "2\n",
      "2\n",
      "2\n",
      "2\n",
      "2\n",
      "2\n",
      "2\n",
      "2\n",
      "2\n",
      "8\n",
      "2\n",
      "2\n",
      "2\n",
      "1\n",
      "1\n",
      "1\n",
      "2\n",
      "2\n",
      "1\n",
      "2\n",
      "2\n",
      "7\n",
      "2\n",
      "8\n",
      "2\n",
      "2\n",
      "1\n",
      "2\n",
      "2\n",
      "2\n",
      "2\n",
      "2\n",
      "2\n",
      "8\n",
      "2\n",
      "2\n",
      "2\n",
      "2\n",
      "2\n",
      "2\n",
      "2\n",
      "# the current counter is 420\n",
      "2\n",
      "2\n",
      "2\n",
      "2\n",
      "2\n",
      "2\n",
      "2\n",
      "2\n",
      "2\n",
      "2\n",
      "1\n",
      "2\n",
      "1\n",
      "2\n",
      "1\n",
      "2\n",
      "2\n",
      "0\n",
      "2\n",
      "8\n",
      "8\n",
      "8\n",
      "2\n",
      "2\n",
      "2\n",
      "0\n",
      "1\n",
      "2\n",
      "1\n",
      "2\n",
      "1\n",
      "1\n",
      "2\n",
      "1\n",
      "0\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "2\n",
      "2\n",
      "8\n",
      "2\n",
      "0\n",
      "0\n",
      "1\n",
      "2\n",
      "2\n",
      "2\n",
      "2\n",
      "1\n",
      "2\n",
      "2\n",
      "2\n",
      "0\n",
      "2\n",
      "2\n",
      "8\n",
      "8\n",
      "2\n",
      "# the current counter is 480\n",
      "2\n",
      "8\n",
      "8\n",
      "2\n",
      "2\n",
      "1\n",
      "1\n",
      "0\n",
      "2\n",
      "2\n",
      "0\n",
      "2\n",
      "2\n",
      "1\n",
      "2\n",
      "2\n",
      "0\n",
      "2\n",
      "2\n",
      "1\n",
      "2\n",
      "2\n",
      "1\n",
      "2\n",
      "2\n",
      "2\n",
      "1\n",
      "1\n",
      "2\n",
      "1\n",
      "2\n",
      "1\n",
      "2\n",
      "1\n",
      "2\n",
      "1\n",
      "2\n",
      "2\n",
      "2\n",
      "2\n",
      "2\n",
      "2\n",
      "2\n",
      "2\n",
      "2\n",
      "2\n",
      "2\n",
      "1\n",
      "1\n",
      "2\n",
      "1\n",
      "2\n",
      "2\n",
      "2\n",
      "2\n",
      "1\n",
      "2\n",
      "2\n",
      "2\n",
      "2\n",
      "# the current counter is 540\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "2\n",
      "2\n",
      "2\n",
      "1\n",
      "8\n",
      "1\n",
      "2\n",
      "8\n",
      "8\n",
      "8\n",
      "1\n",
      "2\n",
      "2\n",
      "0\n",
      "1\n",
      "2\n",
      "1\n",
      "1\n",
      "2\n",
      "2\n",
      "2\n",
      "2\n",
      "2\n",
      "8\n",
      "1\n",
      "1\n",
      "2\n",
      "2\n",
      "2\n",
      "1\n",
      "2\n",
      "2\n",
      "7\n",
      "2\n",
      "1\n",
      "2\n",
      "2\n",
      "0\n",
      "2\n",
      "2\n",
      "2\n",
      "2\n",
      "2\n",
      "1\n",
      "2\n",
      "2\n",
      "8\n",
      "8\n",
      "1\n",
      "1\n",
      "8\n",
      "7\n",
      "2\n",
      "2\n",
      "2\n",
      "# the current counter is 600\n",
      "2\n",
      "2\n",
      "2\n",
      "2\n",
      "2\n",
      "8\n",
      "1\n",
      "2\n",
      "2\n",
      "1\n",
      "1\n",
      "2\n",
      "1\n",
      "1\n",
      "1\n",
      "2\n",
      "2\n",
      "2\n",
      "2\n",
      "1\n",
      "2\n",
      "1\n",
      "2\n",
      "2\n",
      "2\n",
      "0\n",
      "2\n",
      "2\n",
      "2\n",
      "1\n",
      "2\n",
      "2\n",
      "1\n",
      "1\n",
      "2\n",
      "2\n",
      "2\n",
      "1\n",
      "2\n",
      "10\n",
      "2\n",
      "8\n",
      "2\n",
      "1\n",
      "2\n",
      "2\n",
      "2\n",
      "7\n",
      "3\n",
      "2\n",
      "2\n",
      "2\n",
      "2\n",
      "2\n",
      "7\n",
      "2\n",
      "2\n",
      "2\n",
      "2\n",
      "2\n",
      "# the current counter is 660\n",
      "2\n",
      "2\n",
      "2\n",
      "2\n",
      "2\n",
      "2\n",
      "2\n",
      "2\n",
      "1\n",
      "2\n",
      "8\n",
      "2\n",
      "2\n",
      "2\n",
      "2\n",
      "2\n",
      "2\n",
      "1\n",
      "2\n",
      "7\n",
      "2\n",
      "1\n",
      "2\n",
      "2\n",
      "2\n",
      "1\n",
      "0\n",
      "2\n",
      "2\n",
      "1\n",
      "2\n",
      "1\n",
      "2\n",
      "0\n",
      "8\n",
      "2\n",
      "2\n",
      "2\n",
      "2\n",
      "2\n",
      "1\n",
      "2\n",
      "1\n",
      "2\n",
      "2\n",
      "2\n",
      "2\n",
      "2\n",
      "2\n",
      "2\n",
      "2\n",
      "7\n",
      "3\n",
      "2\n",
      "2\n",
      "2\n",
      "2\n",
      "2\n",
      "2\n",
      "1\n",
      "# the current counter is 720\n",
      "8\n",
      "8\n",
      "2\n",
      "2\n",
      "2\n",
      "2\n",
      "8\n",
      "1\n",
      "2\n",
      "0\n",
      "2\n",
      "2\n",
      "2\n",
      "2\n",
      "2\n",
      "2\n",
      "2\n",
      "2\n",
      "2\n",
      "2\n",
      "2\n",
      "8\n",
      "2\n",
      "8\n",
      "2\n",
      "0\n",
      "2\n",
      "0\n",
      "1\n",
      "2\n",
      "2\n",
      "2\n",
      "2\n",
      "2\n",
      "2\n",
      "1\n",
      "2\n",
      "1\n",
      "2\n",
      "0\n",
      "2\n",
      "1\n",
      "1\n",
      "2\n",
      "0\n",
      "2\n",
      "0\n",
      "2\n",
      "2\n",
      "1\n",
      "0\n",
      "2\n",
      "2\n",
      "0\n",
      "2\n",
      "0\n",
      "1\n",
      "2\n",
      "2\n",
      "2\n",
      "# the current counter is 780\n",
      "2\n",
      "2\n",
      "1\n",
      "2\n",
      "1\n",
      "2\n",
      "2\n",
      "2\n",
      "2\n",
      "2\n",
      "2\n",
      "2\n",
      "2\n",
      "2\n",
      "1\n",
      "2\n",
      "2\n",
      "2\n",
      "8\n",
      "8\n",
      "2\n",
      "2\n",
      "1\n",
      "2\n",
      "2\n",
      "2\n",
      "2\n",
      "2\n",
      "2\n",
      "0\n",
      "1\n",
      "2\n",
      "2\n",
      "2\n",
      "2\n",
      "2\n",
      "2\n",
      "1\n",
      "1\n",
      "2\n",
      "2\n",
      "8\n",
      "8\n",
      "2\n",
      "1\n",
      "2\n",
      "2\n",
      "1\n",
      "2\n",
      "8\n",
      "8\n",
      "8\n",
      "1\n",
      "1\n",
      "1\n",
      "2\n",
      "2\n",
      "2\n",
      "1\n",
      "2\n",
      "# the current counter is 840\n",
      "0\n",
      "2\n",
      "2\n",
      "2\n",
      "1\n",
      "2\n",
      "2\n",
      "2\n",
      "1\n",
      "1\n",
      "2\n",
      "2\n",
      "2\n",
      "1\n",
      "2\n",
      "8\n",
      "2\n",
      "2\n",
      "2\n",
      "2\n",
      "2\n",
      "2\n",
      "2\n",
      "8\n",
      "0\n",
      "2\n",
      "2\n",
      "2\n",
      "2\n",
      "2\n",
      "2\n",
      "2\n",
      "2\n",
      "2\n",
      "2\n",
      "2\n",
      "1\n",
      "2\n",
      "2\n",
      "2\n",
      "7\n",
      "2\n",
      "2\n",
      "2\n",
      "7\n",
      "2\n",
      "2\n",
      "2\n",
      "2\n",
      "2\n",
      "2\n",
      "2\n",
      "2\n",
      "2\n",
      "2\n",
      "0\n",
      "2\n",
      "2\n",
      "0\n",
      "2\n",
      "# the current counter is 900\n",
      "8\n",
      "2\n",
      "8\n",
      "8\n",
      "2\n",
      "0\n",
      "2\n",
      "2\n",
      "2\n",
      "2\n",
      "1\n",
      "0\n",
      "2\n",
      "0\n",
      "8\n",
      "1\n",
      "1\n",
      "2\n",
      "2\n",
      "2\n",
      "2\n",
      "2\n",
      "2\n",
      "1\n",
      "2\n",
      "0\n",
      "8\n",
      "2\n",
      "2\n",
      "2\n",
      "0\n",
      "2\n",
      "2\n",
      "2\n",
      "2\n",
      "2\n",
      "2\n",
      "2\n",
      "1\n",
      "2\n",
      "1\n",
      "2\n",
      "2\n",
      "2\n",
      "2\n",
      "2\n",
      "2\n",
      "2\n",
      "2\n",
      "8\n",
      "2\n",
      "2\n",
      "8\n",
      "2\n",
      "2\n",
      "2\n",
      "8\n",
      "2\n",
      "2\n",
      "1\n",
      "# the current counter is 960\n",
      "2\n",
      "2\n",
      "2\n",
      "2\n",
      "2\n",
      "2\n",
      "1\n",
      "0\n",
      "2\n",
      "8\n",
      "2\n",
      "2\n",
      "1\n",
      "2\n",
      "2\n",
      "1\n",
      "1\n",
      "10\n",
      "1\n",
      "2\n",
      "2\n",
      "2\n",
      "2\n",
      "2\n",
      "2\n",
      "2\n",
      "0\n",
      "0\n",
      "1\n",
      "2\n",
      "2\n",
      "2\n",
      "1\n",
      "0\n",
      "1\n",
      "1\n",
      "2\n",
      "2\n",
      "1\n",
      "1\n",
      "2\n",
      "1\n",
      "2\n",
      "1\n",
      "2\n",
      "10\n",
      "2\n",
      "8\n",
      "2\n",
      "2\n",
      "2\n",
      "0\n",
      "1\n",
      "1\n",
      "8\n",
      "8\n",
      "2\n",
      "2\n",
      "2\n",
      "2\n",
      "# the current counter is 1020\n",
      "2\n",
      "1\n",
      "1\n",
      "0\n",
      "2\n",
      "2\n",
      "0\n",
      "2\n",
      "2\n",
      "2\n",
      "7\n",
      "8\n",
      "1\n",
      "2\n",
      "2\n",
      "0\n",
      "2\n",
      "2\n",
      "1\n",
      "2\n",
      "1\n",
      "0\n",
      "0\n",
      "1\n",
      "1\n",
      "2\n",
      "1\n",
      "0\n",
      "2\n",
      "2\n",
      "1\n",
      "1\n",
      "1\n",
      "8\n",
      "1\n",
      "2\n",
      "2\n",
      "1\n",
      "2\n",
      "2\n",
      "2\n",
      "2\n",
      "1\n",
      "2\n",
      "1\n",
      "2\n",
      "1\n",
      "1\n",
      "2\n",
      "1\n",
      "1\n",
      "2\n",
      "all 1071 prompts took: 2752.8026034832 seconds\n"
     ]
    }
   ],
   "source": [
    "# if the column gpt_3_5_ratings already exists, then read skip this execution\n",
    "if \"gpt_3_5_ratings\" in list(experiment_df.columns):\n",
    "    print(\"gpt3_5_ratings already exists\")\n",
    "else:\n",
    "    # get the score for each prompt\n",
    "    gpt_3_5_responses = []\n",
    "    gpt_3_5_response_objects = []\n",
    "    start_timestamp = time.time()\n",
    "    last_start_timestamp = time.time()\n",
    "    counter = 0\n",
    "    for prompt in verses:\n",
    "        counter += 1\n",
    "        if counter%60 == 0:\n",
    "            print(\"# the current counter is\", counter)\n",
    "        response = get_GPT3_5_response(prompt)\n",
    "        print(response.choices[0].message.content)\n",
    "        gpt_3_5_responses.append(response.choices[0].message.content)\n",
    "        gpt_3_5_response_objects.append(response)\n",
    "    \n",
    "    total_time_for_gpt3_5 = time.time() - start_timestamp\n",
    "    print(\"all\", len(verses), \"prompts took:\", total_time_for_gpt3_5, \"seconds\")\n",
    "    \n",
    "    gpt_3_5_responses_int = [convert_to_int(response.strip()) if response.strip() != \"\" else -3 for response in gpt_3_5_responses ]\n",
    "\n",
    "    experiment_df.insert(2,\"gpt_3_5_ratings\", gpt_3_5_responses_int)\n",
    "    experiment_df.insert(2,\"gpt_3_5_response_objects\", gpt_3_5_response_objects)\n",
    "    \n",
    "    # save time in times_df\n",
    "    times_df = pd.concat([times_df, pd.DataFrame([{'name': 'gpt_3_5', 'sec': total_time_for_gpt3_5}])])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 48,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "ChatCompletionMessage(content='0', role='assistant', function_call=None, tool_calls=None)\n"
     ]
    }
   ],
   "source": [
    "from openai import OpenAI\n",
    "\n",
    "client = OpenAI()\n",
    "def get_GPT4_response(prompt):\n",
    "  completion = client.chat.completions.create(\n",
    "    model=\"gpt-4\",\n",
    "    max_tokens=2,\n",
    "    temperature=0,\n",
    "    messages=[\n",
    "      {\"role\": \"system\", \"content\": system_prompt},\n",
    "      {\"role\": \"user\", \"content\": prompt}\n",
    "    ]\n",
    "  )\n",
    "  return completion\n",
    "print(get_GPT4_response(verses[30]).choices[0].message)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 49,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "1\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "1\n",
      "1\n",
      "1\n",
      "0\n",
      "0\n",
      "0\n",
      "5\n",
      "0\n",
      "1\n",
      "1\n",
      "1\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "1\n",
      "1\n",
      "0\n",
      "1\n",
      "1\n",
      "4\n",
      "2\n",
      "0\n",
      "1\n",
      "0\n",
      "8\n",
      "0\n",
      "2\n",
      "1\n",
      "2\n",
      "0\n",
      "2\n",
      "4\n",
      "2\n",
      "7\n",
      "1\n",
      "2\n",
      "1\n",
      "4\n",
      "0\n",
      "1\n",
      "2\n",
      "# the current counter is 60\n",
      "1\n",
      "4\n",
      "7\n",
      "1\n",
      "1\n",
      "0\n",
      "8\n",
      "8\n",
      "1\n",
      "2\n",
      "1\n",
      "7\n",
      "1\n",
      "1\n",
      "7\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "0\n",
      "0\n",
      "2\n",
      "4\n",
      "0\n",
      "5\n",
      "7\n",
      "1\n",
      "7\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "0\n",
      "7\n",
      "7\n",
      "1\n",
      "7\n",
      "0\n",
      "0\n",
      "0\n",
      "8\n",
      "9\n",
      "7\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "0\n",
      "0\n",
      "1\n",
      "6\n",
      "0\n",
      "0\n",
      "1\n",
      "2\n",
      "2\n",
      "1\n",
      "0\n",
      "1\n",
      "9\n",
      "# the current counter is 120\n",
      "9\n",
      "0\n",
      "1\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "1\n",
      "1\n",
      "9\n",
      "8\n",
      "8\n",
      "1\n",
      "1\n",
      "8\n",
      "1\n",
      "4\n",
      "2\n",
      "5\n",
      "2\n",
      "2\n",
      "2\n",
      "2\n",
      "2\n",
      "1\n",
      "1\n",
      "1\n",
      "0\n",
      "1\n",
      "1\n",
      "0\n",
      "4\n",
      "0\n",
      "1\n",
      "7\n",
      "7\n",
      "2\n",
      "7\n",
      "1\n",
      "2\n",
      "0\n",
      "2\n",
      "2\n",
      "7\n",
      "1\n",
      "2\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "2\n",
      "2\n",
      "0\n",
      "1\n",
      "2\n",
      "2\n",
      "7\n",
      "1\n",
      "1\n",
      "# the current counter is 180\n",
      "7\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "8\n",
      "8\n",
      "1\n",
      "1\n",
      "0\n",
      "0\n",
      "1\n",
      "0\n",
      "2\n",
      "1\n",
      "1\n",
      "1\n",
      "2\n",
      "7\n",
      "2\n",
      "0\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "0\n",
      "2\n",
      "0\n",
      "1\n",
      "1\n",
      "0\n",
      "0\n",
      "8\n",
      "0\n",
      "1\n",
      "1\n",
      "1\n",
      "6\n",
      "2\n",
      "7\n",
      "8\n",
      "1\n",
      "7\n",
      "1\n",
      "2\n",
      "2\n",
      "2\n",
      "0\n",
      "2\n",
      "7\n",
      "0\n",
      "1\n",
      "1\n",
      "0\n",
      "1\n",
      "0\n",
      "1\n",
      "0\n",
      "1\n",
      "# the current counter is 240\n",
      "1\n",
      "1\n",
      "0\n",
      "0\n",
      "7\n",
      "1\n",
      "7\n",
      "2\n",
      "2\n",
      "7\n",
      "7\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "7\n",
      "7\n",
      "1\n",
      "0\n",
      "1\n",
      "1\n",
      "0\n",
      "1\n",
      "1\n",
      "2\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "0\n",
      "2\n",
      "1\n",
      "7\n",
      "6\n",
      "2\n",
      "0\n",
      "0\n",
      "0\n",
      "4\n",
      "7\n",
      "2\n",
      "2\n",
      "8\n",
      "8\n",
      "1\n",
      "0\n",
      "1\n",
      "7\n",
      "1\n",
      "7\n",
      "8\n",
      "8\n",
      "4\n",
      "0\n",
      "2\n",
      "8\n",
      "2\n",
      "1\n",
      "7\n",
      "1\n",
      "# the current counter is 300\n",
      "2\n",
      "2\n",
      "0\n",
      "0\n",
      "0\n",
      "1\n",
      "1\n",
      "8\n",
      "8\n",
      "5\n",
      "7\n",
      "9\n",
      "8\n",
      "0\n",
      "1\n",
      "1\n",
      "1\n",
      "2\n",
      "1\n",
      "0\n",
      "1\n",
      "1\n",
      "7\n",
      "7\n",
      "2\n",
      "1\n",
      "1\n",
      "7\n",
      "0\n",
      "0\n",
      "1\n",
      "1\n",
      "1\n",
      "7\n",
      "4\n",
      "1\n",
      "2\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "0\n",
      "0\n",
      "7\n",
      "2\n",
      "2\n",
      "2\n",
      "1\n",
      "3\n",
      "1\n",
      "1\n",
      "0\n",
      "1\n",
      "0\n",
      "0\n",
      "1\n",
      "1\n",
      "0\n",
      "1\n",
      "2\n",
      "# the current counter is 360\n",
      "1\n",
      "0\n",
      "0\n",
      "1\n",
      "1\n",
      "1\n",
      "0\n",
      "1\n",
      "0\n",
      "0\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "2\n",
      "1\n",
      "0\n",
      "0\n",
      "1\n",
      "1\n",
      "0\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "7\n",
      "1\n",
      "1\n",
      "7\n",
      "1\n",
      "2\n",
      "0\n",
      "0\n",
      "1\n",
      "0\n",
      "1\n",
      "0\n",
      "0\n",
      "1\n",
      "1\n",
      "3\n",
      "2\n",
      "2\n",
      "0\n",
      "1\n",
      "1\n",
      "1\n",
      "2\n",
      "2\n",
      "1\n",
      "2\n",
      "1\n",
      "2\n",
      "1\n",
      "1\n",
      "2\n",
      "7\n",
      "4\n",
      "1\n",
      "1\n",
      "# the current counter is 420\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "0\n",
      "0\n",
      "1\n",
      "2\n",
      "7\n",
      "0\n",
      "7\n",
      "1\n",
      "8\n",
      "0\n",
      "1\n",
      "1\n",
      "7\n",
      "0\n",
      "1\n",
      "0\n",
      "1\n",
      "0\n",
      "0\n",
      "2\n",
      "1\n",
      "0\n",
      "0\n",
      "1\n",
      "0\n",
      "1\n",
      "1\n",
      "0\n",
      "1\n",
      "4\n",
      "0\n",
      "1\n",
      "0\n",
      "2\n",
      "1\n",
      "2\n",
      "2\n",
      "1\n",
      "0\n",
      "1\n",
      "0\n",
      "0\n",
      "2\n",
      "7\n",
      "2\n",
      "1\n",
      "2\n",
      "# the current counter is 480\n",
      "1\n",
      "8\n",
      "8\n",
      "4\n",
      "2\n",
      "0\n",
      "0\n",
      "0\n",
      "1\n",
      "1\n",
      "0\n",
      "1\n",
      "1\n",
      "0\n",
      "1\n",
      "1\n",
      "0\n",
      "2\n",
      "1\n",
      "0\n",
      "1\n",
      "1\n",
      "1\n",
      "2\n",
      "0\n",
      "0\n",
      "0\n",
      "1\n",
      "1\n",
      "1\n",
      "2\n",
      "2\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "2\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "0\n",
      "1\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "1\n",
      "0\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "0\n",
      "0\n",
      "1\n",
      "# the current counter is 540\n",
      "1\n",
      "1\n",
      "0\n",
      "0\n",
      "0\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "9\n",
      "1\n",
      "2\n",
      "9\n",
      "8\n",
      "2\n",
      "1\n",
      "1\n",
      "2\n",
      "0\n",
      "0\n",
      "1\n",
      "0\n",
      "6\n",
      "2\n",
      "0\n",
      "1\n",
      "0\n",
      "0\n",
      "7\n",
      "0\n",
      "0\n",
      "2\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "2\n",
      "7\n",
      "7\n",
      "2\n",
      "0\n",
      "0\n",
      "0\n",
      "1\n",
      "1\n",
      "0\n",
      "2\n",
      "4\n",
      "1\n",
      "2\n",
      "4\n",
      "9\n",
      "9\n",
      "0\n",
      "1\n",
      "8\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "# the current counter is 600\n",
      "1\n",
      "1\n",
      "0\n",
      "0\n",
      "1\n",
      "1\n",
      "1\n",
      "0\n",
      "1\n",
      "2\n",
      "0\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "7\n",
      "1\n",
      "0\n",
      "0\n",
      "1\n",
      "0\n",
      "1\n",
      "1\n",
      "0\n",
      "2\n",
      "1\n",
      "2\n",
      "1\n",
      "8\n",
      "1\n",
      "0\n",
      "0\n",
      "1\n",
      "1\n",
      "0\n",
      "0\n",
      "1\n",
      "8\n",
      "2\n",
      "7\n",
      "7\n",
      "1\n",
      "1\n",
      "8\n",
      "1\n",
      "8\n",
      "4\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "0\n",
      "2\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "# the current counter is 660\n",
      "7\n",
      "1\n",
      "1\n",
      "1\n",
      "2\n",
      "5\n",
      "7\n",
      "8\n",
      "1\n",
      "1\n",
      "8\n",
      "7\n",
      "1\n",
      "1\n",
      "7\n",
      "8\n",
      "8\n",
      "1\n",
      "1\n",
      "2\n",
      "1\n",
      "1\n",
      "1\n",
      "0\n",
      "1\n",
      "1\n",
      "0\n",
      "2\n",
      "1\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "2\n",
      "1\n",
      "1\n",
      "1\n",
      "0\n",
      "1\n",
      "1\n",
      "1\n",
      "0\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "0\n",
      "2\n",
      "4\n",
      "0\n",
      "2\n",
      "1\n",
      "1\n",
      "1\n",
      "2\n",
      "1\n",
      "0\n",
      "# the current counter is 720\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "4\n",
      "1\n",
      "7\n",
      "0\n",
      "1\n",
      "0\n",
      "0\n",
      "1\n",
      "0\n",
      "1\n",
      "2\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "7\n",
      "1\n",
      "1\n",
      "0\n",
      "1\n",
      "1\n",
      "0\n",
      "0\n",
      "1\n",
      "0\n",
      "0\n",
      "1\n",
      "1\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "1\n",
      "0\n",
      "1\n",
      "0\n",
      "1\n",
      "0\n",
      "1\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "1\n",
      "0\n",
      "1\n",
      "0\n",
      "0\n",
      "1\n",
      "8\n",
      "1\n",
      "# the current counter is 780\n",
      "1\n",
      "1\n",
      "0\n",
      "0\n",
      "1\n",
      "6\n",
      "7\n",
      "1\n",
      "1\n",
      "6\n",
      "1\n",
      "1\n",
      "0\n",
      "1\n",
      "0\n",
      "0\n",
      "0\n",
      "1\n",
      "4\n",
      "2\n",
      "2\n",
      "2\n",
      "2\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "8\n",
      "1\n",
      "0\n",
      "1\n",
      "2\n",
      "0\n",
      "1\n",
      "2\n",
      "1\n",
      "1\n",
      "1\n",
      "3\n",
      "2\n",
      "5\n",
      "9\n",
      "2\n",
      "1\n",
      "1\n",
      "7\n",
      "1\n",
      "1\n",
      "7\n",
      "2\n",
      "7\n",
      "2\n",
      "1\n",
      "7\n",
      "1\n",
      "1\n",
      "1\n",
      "0\n",
      "1\n",
      "# the current counter is 840\n",
      "0\n",
      "0\n",
      "2\n",
      "1\n",
      "0\n",
      "0\n",
      "1\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "0\n",
      "1\n",
      "0\n",
      "1\n",
      "1\n",
      "1\n",
      "7\n",
      "1\n",
      "0\n",
      "2\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "0\n",
      "1\n",
      "1\n",
      "1\n",
      "5\n",
      "0\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "2\n",
      "2\n",
      "2\n",
      "1\n",
      "2\n",
      "1\n",
      "7\n",
      "0\n",
      "1\n",
      "0\n",
      "0\n",
      "1\n",
      "# the current counter is 900\n",
      "7\n",
      "1\n",
      "7\n",
      "2\n",
      "2\n",
      "1\n",
      "2\n",
      "7\n",
      "2\n",
      "2\n",
      "1\n",
      "0\n",
      "8\n",
      "0\n",
      "1\n",
      "1\n",
      "2\n",
      "1\n",
      "2\n",
      "1\n",
      "1\n",
      "1\n",
      "2\n",
      "1\n",
      "1\n",
      "1\n",
      "2\n",
      "1\n",
      "1\n",
      "1\n",
      "0\n",
      "2\n",
      "2\n",
      "1\n",
      "2\n",
      "0\n",
      "1\n",
      "1\n",
      "1\n",
      "2\n",
      "1\n",
      "7\n",
      "1\n",
      "7\n",
      "1\n",
      "8\n",
      "7\n",
      "7\n",
      "7\n",
      "8\n",
      "2\n",
      "7\n",
      "9\n",
      "1\n",
      "2\n",
      "2\n",
      "2\n",
      "2\n",
      "1\n",
      "0\n",
      "# the current counter is 960\n",
      "2\n",
      "1\n",
      "2\n",
      "1\n",
      "1\n",
      "1\n",
      "2\n",
      "1\n",
      "7\n",
      "1\n",
      "0\n",
      "1\n",
      "0\n",
      "1\n",
      "0\n",
      "1\n",
      "0\n",
      "8\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "0\n",
      "1\n",
      "1\n",
      "2\n",
      "1\n",
      "1\n",
      "2\n",
      "1\n",
      "1\n",
      "0\n",
      "1\n",
      "0\n",
      "1\n",
      "0\n",
      "1\n",
      "1\n",
      "1\n",
      "0\n",
      "0\n",
      "0\n",
      "1\n",
      "0\n",
      "1\n",
      "1\n",
      "1\n",
      "7\n",
      "2\n",
      "2\n",
      "2\n",
      "7\n",
      "2\n",
      "2\n",
      "8\n",
      "7\n",
      "8\n",
      "7\n",
      "1\n",
      "6\n",
      "# the current counter is 1020\n",
      "1\n",
      "0\n",
      "1\n",
      "1\n",
      "0\n",
      "1\n",
      "0\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "9\n",
      "0\n",
      "2\n",
      "1\n",
      "7\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "1\n",
      "0\n",
      "0\n",
      "1\n",
      "0\n",
      "1\n",
      "1\n",
      "0\n",
      "0\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "0\n",
      "2\n",
      "1\n",
      "1\n",
      "1\n",
      "2\n",
      "1\n",
      "1\n",
      "1\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "2\n",
      "1\n",
      "0\n",
      "1\n",
      "1\n",
      "all 1071 prompts took: 1070.9602489471436 seconds\n"
     ]
    }
   ],
   "source": [
    "# if the column gpt_4_ratings already exists, then read skip this execution\n",
    "if \"gpt_4_ratings\" in list(experiment_df.columns):\n",
    "    print(\"gpt_4_ratings already exists\")\n",
    "else:\n",
    "    # get the score for each prompt\n",
    "    gpt_4_responses = []\n",
    "    gpt_4_response_objects = []\n",
    "    start_timestamp = time.time()\n",
    "    last_start_timestamp = time.time()\n",
    "    counter = 0\n",
    "    for prompt in verses:\n",
    "        counter += 1\n",
    "        if counter%60 == 0:\n",
    "            print(\"# the current counter is\", counter)\n",
    "        response = get_GPT4_response(prompt)\n",
    "        print(response.choices[0].message.content)\n",
    "        gpt_4_responses.append(response.choices[0].message.content)\n",
    "        gpt_4_response_objects.append(response)\n",
    "    \n",
    "    total_time_for_gpt_4 = time.time() - start_timestamp\n",
    "    print(\"all\", len(verses), \"prompts took:\", total_time_for_gpt_4, \"seconds\")\n",
    "    \n",
    "    gpt_4_responses_int = [convert_to_int(response.strip()) if response.strip() != \"\" else -3 for response in gpt_4_responses ]\n",
    "\n",
    "    experiment_df.insert(2,\"gpt_4_ratings\", gpt_4_responses_int)\n",
    "    experiment_df.insert(2,\"gpt_4_response_objects\", gpt_4_response_objects)\n",
    "    \n",
    "    # save time in times_df\n",
    "    times_df = pd.concat([times_df, pd.DataFrame([{'name': 'gpt_4', 'sec': total_time_for_gpt_4}])])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 51,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Index(['vers_text', 'bible_vers_id', 'gpt_4_response_objects', 'gpt_4_ratings',\n",
       "       'mistral_medium_response_objects', 'mistral_medium_ratings',\n",
       "       'mistral_small_response_objects', 'mistral_small_ratings',\n",
       "       'mistral_tiny_response_objects', 'mistral_tiny_ratings',\n",
       "       'ollama_llama13B_response_objects', 'ollama_llama13B_ratings',\n",
       "       'ollama_llama7B_response_objects', 'ollama_llama7B_ratings',\n",
       "       'ollama_mistral7B_response_objects', 'ollama_mistral7B_ratings',\n",
       "       'gpt_3_5_response_objects', 'gpt_3_5_ratings',\n",
       "       'gemini_pro_response_objects', 'gemini_pro_ratings',\n",
       "       'text_bison_response_objects', 'text_bison_ratings', 'vers_references'],\n",
       "      dtype='object')"
      ]
     },
     "execution_count": 51,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "experiment_df.columns"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 57,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>vers_text</th>\n",
       "      <th>gemini_pro_ratings</th>\n",
       "      <th>text_bison_ratings</th>\n",
       "      <th>gpt_3_5_ratings</th>\n",
       "      <th>gpt_4_ratings</th>\n",
       "      <th>ollama_mistral7B_ratings</th>\n",
       "      <th>mistral_medium_ratings</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>551</th>\n",
       "      <td>Then Jesus said to his disciples, \"If anyone desires to come after me, let him deny himself, and take up his cross, and follow me.</td>\n",
       "      <td>8</td>\n",
       "      <td>8</td>\n",
       "      <td>8</td>\n",
       "      <td>9</td>\n",
       "      <td>10</td>\n",
       "      <td>9</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>133</th>\n",
       "      <td>But I tell you, love your enemies, bless those who curse you, do good to those who hate you, and pray for those who mistreat you and persecute you,</td>\n",
       "      <td>7</td>\n",
       "      <td>5</td>\n",
       "      <td>2</td>\n",
       "      <td>8</td>\n",
       "      <td>10</td>\n",
       "      <td>8</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>548</th>\n",
       "      <td>From that time, Jesus began to show his disciples that he must go to Jerusalem and suffer many things from the elders, chief priests, and scribes, and be killed, and the third day be raised up.</td>\n",
       "      <td>7</td>\n",
       "      <td>2</td>\n",
       "      <td>8</td>\n",
       "      <td>9</td>\n",
       "      <td>10</td>\n",
       "      <td>8</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>119</th>\n",
       "      <td>If your right hand causes you to stumble, cut it off, and throw it away from you: for it is profitable for you that one of your members should perish, and not your whole body be thrown into Gehenna.</td>\n",
       "      <td>8</td>\n",
       "      <td>5</td>\n",
       "      <td>8</td>\n",
       "      <td>9</td>\n",
       "      <td>0</td>\n",
       "      <td>8</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>220</th>\n",
       "      <td>Jesus said to him, \"The foxes have holes, and the birds of the sky have nests, but the Son of Man has nowhere to lay his head.\"</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>8</td>\n",
       "      <td>8</td>\n",
       "      <td>10</td>\n",
       "      <td>8</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>666</th>\n",
       "      <td>and will hand him over to the Gentiles to mock, to scourge, and to crucify; and the third day he will be raised up.\"</td>\n",
       "      <td>5</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>8</td>\n",
       "      <td>0</td>\n",
       "      <td>8</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>310</th>\n",
       "      <td>He who doesn't take his cross and follow after me, isn't worthy of me.</td>\n",
       "      <td>7</td>\n",
       "      <td>5</td>\n",
       "      <td>8</td>\n",
       "      <td>9</td>\n",
       "      <td>5</td>\n",
       "      <td>8</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>311</th>\n",
       "      <td>He who finds his life will lose it; and he who loses his life for my sake will find it.</td>\n",
       "      <td>8</td>\n",
       "      <td>5</td>\n",
       "      <td>8</td>\n",
       "      <td>8</td>\n",
       "      <td>5</td>\n",
       "      <td>8</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>130</th>\n",
       "      <td>Whoever compels you to go one mile, go with him two.</td>\n",
       "      <td>5</td>\n",
       "      <td>5</td>\n",
       "      <td>7</td>\n",
       "      <td>8</td>\n",
       "      <td>0</td>\n",
       "      <td>8</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>552</th>\n",
       "      <td>For whoever desires to save his life will lose it, and whoever will lose his life for my sake will find it.</td>\n",
       "      <td>7</td>\n",
       "      <td>0</td>\n",
       "      <td>8</td>\n",
       "      <td>8</td>\n",
       "      <td>5</td>\n",
       "      <td>8</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>946</th>\n",
       "      <td>He took with him Peter and the two sons of Zebedee, and began to be sorrowful and severely troubled.</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>7</td>\n",
       "      <td>5</td>\n",
       "      <td>8</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1010</th>\n",
       "      <td>Then he released to them Barabbas, but Jesus he flogged and delivered to be crucified.</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>7</td>\n",
       "      <td>10</td>\n",
       "      <td>8</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>777</th>\n",
       "      <td>For they bind heavy burdens that are grievous to be borne, and lay them on men's shoulders; but they themselves will not lift a finger to help them.</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>8</td>\n",
       "      <td>10</td>\n",
       "      <td>8</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>384</th>\n",
       "      <td>For as Jonah was three days and three nights in the belly of the whale, so will the Son of Man be three days and three nights in the heart of the earth.</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>7</td>\n",
       "      <td>0</td>\n",
       "      <td>8</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>659</th>\n",
       "      <td>saying, 'These last have spent one hour, and you have made them equal to us, who have borne the burden of the day and the scorching heat!'</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>7</td>\n",
       "      <td>2</td>\n",
       "      <td>8</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>807</th>\n",
       "      <td>Therefore, behold, I send to you prophets, wise men, and scribes. Some of them you will kill and crucify; and some of them you will scourge in your synagogues, and persecute from city to city;</td>\n",
       "      <td>7</td>\n",
       "      <td>5</td>\n",
       "      <td>2</td>\n",
       "      <td>8</td>\n",
       "      <td>10</td>\n",
       "      <td>8</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>646</th>\n",
       "      <td>Everyone who has left houses, or brothers, or sisters, or father, or mother, or wife, or children, or lands, for my name's sake, will receive one hundred times, and will inherit eternal life.</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>7</td>\n",
       "      <td>8</td>\n",
       "      <td>5</td>\n",
       "      <td>8</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>638</th>\n",
       "      <td>Jesus said to him, \"If you want to be perfect, go, sell what you have, and give to the poor, and you will have treasure in heaven; and come, follow me.\"</td>\n",
       "      <td>5</td>\n",
       "      <td>3</td>\n",
       "      <td>10</td>\n",
       "      <td>8</td>\n",
       "      <td>8</td>\n",
       "      <td>8</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>118</th>\n",
       "      <td>If your right eye causes you to stumble, pluck it out and throw it away from you. For it is more profitable for you that one of your members should perish, than for your whole body to be cast into Gehenna.</td>\n",
       "      <td>7</td>\n",
       "      <td>5</td>\n",
       "      <td>8</td>\n",
       "      <td>9</td>\n",
       "      <td>10</td>\n",
       "      <td>8</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>440</th>\n",
       "      <td>who having found one pearl of great price, he went and sold all that he had, and bought it.</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>8</td>\n",
       "      <td>8</td>\n",
       "      <td>0</td>\n",
       "      <td>8</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                                                                                                                                                                                          vers_text  \\\n",
       "551   Then Jesus said to his disciples, \"If anyone desires to come after me, let him deny himself, and take up his cross, and follow me.                                                                              \n",
       "133   But I tell you, love your enemies, bless those who curse you, do good to those who hate you, and pray for those who mistreat you and persecute you,                                                             \n",
       "548   From that time, Jesus began to show his disciples that he must go to Jerusalem and suffer many things from the elders, chief priests, and scribes, and be killed, and the third day be raised up.               \n",
       "119   If your right hand causes you to stumble, cut it off, and throw it away from you: for it is profitable for you that one of your members should perish, and not your whole body be thrown into Gehenna.          \n",
       "220   Jesus said to him, \"The foxes have holes, and the birds of the sky have nests, but the Son of Man has nowhere to lay his head.\"                                                                                 \n",
       "666   and will hand him over to the Gentiles to mock, to scourge, and to crucify; and the third day he will be raised up.\"                                                                                            \n",
       "310   He who doesn't take his cross and follow after me, isn't worthy of me.                                                                                                                                          \n",
       "311   He who finds his life will lose it; and he who loses his life for my sake will find it.                                                                                                                         \n",
       "130   Whoever compels you to go one mile, go with him two.                                                                                                                                                            \n",
       "552   For whoever desires to save his life will lose it, and whoever will lose his life for my sake will find it.                                                                                                     \n",
       "946   He took with him Peter and the two sons of Zebedee, and began to be sorrowful and severely troubled.                                                                                                            \n",
       "1010  Then he released to them Barabbas, but Jesus he flogged and delivered to be crucified.                                                                                                                          \n",
       "777   For they bind heavy burdens that are grievous to be borne, and lay them on men's shoulders; but they themselves will not lift a finger to help them.                                                            \n",
       "384   For as Jonah was three days and three nights in the belly of the whale, so will the Son of Man be three days and three nights in the heart of the earth.                                                        \n",
       "659   saying, 'These last have spent one hour, and you have made them equal to us, who have borne the burden of the day and the scorching heat!'                                                                      \n",
       "807   Therefore, behold, I send to you prophets, wise men, and scribes. Some of them you will kill and crucify; and some of them you will scourge in your synagogues, and persecute from city to city;                \n",
       "646   Everyone who has left houses, or brothers, or sisters, or father, or mother, or wife, or children, or lands, for my name's sake, will receive one hundred times, and will inherit eternal life.                 \n",
       "638   Jesus said to him, \"If you want to be perfect, go, sell what you have, and give to the poor, and you will have treasure in heaven; and come, follow me.\"                                                        \n",
       "118   If your right eye causes you to stumble, pluck it out and throw it away from you. For it is more profitable for you that one of your members should perish, than for your whole body to be cast into Gehenna.   \n",
       "440   who having found one pearl of great price, he went and sold all that he had, and bought it.                                                                                                                     \n",
       "\n",
       "      gemini_pro_ratings  text_bison_ratings  gpt_3_5_ratings  gpt_4_ratings  \\\n",
       "551   8                   8                   8                9               \n",
       "133   7                   5                   2                8               \n",
       "548   7                   2                   8                9               \n",
       "119   8                   5                   8                9               \n",
       "220   2                   0                   8                8               \n",
       "666   5                   0                   2                8               \n",
       "310   7                   5                   8                9               \n",
       "311   8                   5                   8                8               \n",
       "130   5                   5                   7                8               \n",
       "552   7                   0                   8                8               \n",
       "946   0                   0                   2                7               \n",
       "1010  0                   0                   0                7               \n",
       "777   0                   0                   2                8               \n",
       "384   0                   0                   2                7               \n",
       "659   0                   0                   2                7               \n",
       "807   7                   5                   2                8               \n",
       "646   0                   0                   7                8               \n",
       "638   5                   3                   10               8               \n",
       "118   7                   5                   8                9               \n",
       "440   0                   0                   8                8               \n",
       "\n",
       "      ollama_mistral7B_ratings  mistral_medium_ratings  \n",
       "551   10                        9                       \n",
       "133   10                        8                       \n",
       "548   10                        8                       \n",
       "119   0                         8                       \n",
       "220   10                        8                       \n",
       "666   0                         8                       \n",
       "310   5                         8                       \n",
       "311   5                         8                       \n",
       "130   0                         8                       \n",
       "552   5                         8                       \n",
       "946   5                         8                       \n",
       "1010  10                        8                       \n",
       "777   10                        8                       \n",
       "384   0                         8                       \n",
       "659   2                         8                       \n",
       "807   10                        8                       \n",
       "646   5                         8                       \n",
       "638   8                         8                       \n",
       "118   10                        8                       \n",
       "440   0                         8                       "
      ]
     },
     "execution_count": 57,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "experiment_df.sort_values(\"mistral_medium_ratings\", ascending=False)[[\"vers_text\", \"gemini_pro_ratings\", \"text_bison_ratings\", \"gpt_3_5_ratings\", \"gpt_4_ratings\", \"ollama_mistral7B_ratings\", \"mistral_medium_ratings\"]].head(20)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {},
   "outputs": [],
   "source": [
    "# # safe the experiment_df as csv\n",
    "# experiment_df.to_csv(\"experiment_df.csv\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Use local Ollama with mistral7B\n",
    "Ollama is a cli that runs large language model inferenc on your local machine.  \n",
    "It is currently the most popular tool to run open source LLMs locally.  \n",
    "Ollama does not yet run on windows natively, so I use WSL to run it on my machine.  \n",
    "my specs: \n",
    "- GTX 1070 (8GB GDDR5-VRAM) (overclocked to 1911 MHz)\n",
    "- i5 6600K (overclocked to 4.5 GHz)\n",
    "\n",
    "```bash\n",
    "ollama create seeker_mistral -f mistral7B_llm\n",
    "```\n",
    "\n",
    "```bash\n",
    "ollama run seeker_mistral \"Enter through the narrow gate. For wide is the gate and broad is the road that leads to destruction, and many enter through it.\"\n",
    "```\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {},
   "outputs": [],
   "source": [
    "import ollama"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      " 0. This verse\n",
      " 0.\n",
      " 0. This verse\n",
      " 0.\n",
      "\n",
      "\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0.\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0.\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 2.\n",
      " 0. This verse\n",
      " 5.\n",
      " 0. This verse\n",
      " 0.\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 5.\n",
      "\n",
      "\n",
      " 0. This verse\n",
      " 0.\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 10. This\n",
      " 0.\n",
      " 0. This verse\n",
      " 10. This\n",
      " 0. This verse\n",
      " 10.\n",
      " 0.\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 5.\n",
      " 0. This verse\n",
      " 5.\n",
      " 0. This verse\n",
      " 5.\n",
      "\n",
      "\n",
      " 8.\n",
      "\n",
      "\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 8.\n",
      "\n",
      "\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 8.\n",
      "\n",
      "\n",
      " 0. This verse\n",
      "# the current counter is 60\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 10.\n",
      "\n",
      " 10.\n",
      " 0. This verse\n",
      " 3. This verse\n",
      " 0.\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 3.\n",
      "\n",
      "\n",
      " 5. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0.\n",
      " 0.\n",
      " 0. This verse\n",
      " 0.\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 10.\n",
      "\n",
      " 0. There is\n",
      " 0. This verse\n",
      " 0.\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 5.\n",
      "\n",
      "\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 10.\n",
      "\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 5.\n",
      "\n",
      "\n",
      " 5.\n",
      "\n",
      "\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0.\n",
      " 0. This verse\n",
      " 10.\n",
      "\n",
      "# the current counter is 120\n",
      " 0. This verse\n",
      " 0.\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0.\n",
      " 0. This verse\n",
      " 0.\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 5.\n",
      "\n",
      "\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 10. This\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0.\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0.\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 5.\n",
      "\n",
      "\n",
      " 0.\n",
      " 3. This verse\n",
      " 0. This verse\n",
      " 0.\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 8.\n",
      "\n",
      "\n",
      " 5.\n",
      "\n",
      "\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0.\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 5.\n",
      "\n",
      "\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 5.\n",
      "\n",
      "\n",
      "# the current counter is 180\n",
      " 3. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 3. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 5.\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0.\n",
      " 0.\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0.\n",
      " 2.\n",
      "\n",
      "\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0.\n",
      " 10. This\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 10. This\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 5.\n",
      "\n",
      "\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 5.\n",
      "\n",
      "\n",
      " 8.\n",
      "\n",
      "\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      "# the current counter is 240\n",
      " 5.\n",
      "\n",
      "\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 5.\n",
      "\n",
      "\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 5.\n",
      "\n",
      "\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 2.\n",
      "\n",
      "\n",
      " 2.\n",
      "\n",
      "\n",
      " 0. This verse\n",
      " 5.\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 5.\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0.\n",
      " 0.\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 5.\n",
      " 5.\n",
      "\n",
      "\n",
      " 5.\n",
      " 0. This verse\n",
      " 0.\n",
      " 10. (\n",
      " 5.\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      "# the current counter is 300\n",
      " 3. This verse\n",
      " 5.\n",
      "\n",
      "\n",
      " 0.\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 8.\n",
      "\n",
      "\n",
      " 10. This\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 5.\n",
      "\n",
      "\n",
      " 5.\n",
      "\n",
      "\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0.\n",
      " 0. This verse\n",
      " 0.\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 8\n",
      "\n",
      "(\n",
      " 0. This verse\n",
      " 0.\n",
      " 8.\n",
      "\n",
      "\n",
      " 0.\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 8\n",
      "\n",
      "(\n",
      " 7.\n",
      "\n",
      "\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 10. (\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 3. This verse\n",
      " 0. This verse\n",
      " 0.\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      "# the current counter is 360\n",
      " 0.\n",
      " 0.\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0.\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 8. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0.\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 8.\n",
      "\n",
      "\n",
      " 0. This verse\n",
      " 10. This\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0.\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 10. This\n",
      " 5.\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 10. This\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 8.\n",
      "\n",
      "\n",
      " 10. This\n",
      " 0. This verse\n",
      " 0.\n",
      "# the current counter is 420\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0.\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0.\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0.\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 5.\n",
      "\n",
      "\n",
      " 10. This\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 3.\n",
      "\n",
      "\n",
      " 0. This verse\n",
      " 0.\n",
      " 0. This verse\n",
      " 0.\n",
      " 10. This\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0.\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0.\n",
      " 0.\n",
      " 0. This verse\n",
      " 5.\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0.\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 5.\n",
      "\n",
      "\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 8.\n",
      "\n",
      "\n",
      " 0. This verse\n",
      " 0. This verse\n",
      "# the current counter is 480\n",
      " 0. This verse\n",
      " 3. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0.\n",
      " 0. This verse\n",
      " 0.\n",
      " 0.\n",
      " 0. This verse\n",
      " 0.\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 5.\n",
      "\n",
      "\n",
      " 0. This verse\n",
      " 0.\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 8\n",
      "\n",
      "(\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0.\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0.\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0.\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 5. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0.\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0.\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0.\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      "# the current counter is 540\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0.\n",
      " 0.\n",
      "\n",
      "\n",
      " 0.\n",
      " 0. This verse\n",
      " 0.\n",
      " 10.\n",
      " 0. This verse\n",
      " 2.\n",
      " 10.\n",
      " 5.\n",
      "\n",
      "\n",
      " 10. This\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0.\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0.\n",
      " 0. This verse\n",
      " 0.\n",
      " 0.\n",
      " 10. This\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 6.\n",
      " 0.\n",
      " 0.\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0.\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0.\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0.\n",
      " 10. This\n",
      " 5.\n",
      "\n",
      "\n",
      " 10. This\n",
      " 10. This\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 8.\n",
      "\n",
      "\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      "# the current counter is 600\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0.\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 10. This\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 10.\n",
      "\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 10. This\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0.\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 5.\n",
      "\n",
      "\n",
      " 8.\n",
      "\n",
      "\n",
      " 5.\n",
      " 5.\n",
      "\n",
      "\n",
      " 10. This\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 5.\n",
      "\n",
      "\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0.\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0.\n",
      "# the current counter is 660\n",
      " 2. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0.\n",
      " 10. This\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0.\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0.\n",
      " 0. This verse\n",
      " 0.\n",
      " 0.\n",
      " 0. This verse\n",
      " 0.\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 8.\n",
      "\n",
      "\n",
      " 5.\n",
      "\n",
      "\n",
      " 0. This verse\n",
      " 0.\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0.\n",
      " 0. This verse\n",
      " 5.\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 10. (\n",
      " 0. This verse\n",
      " 0. This verse\n",
      "# the current counter is 720\n",
      " 10. This\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0.\n",
      " 0.\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0.\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0.\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 5.\n",
      " 0. This verse\n",
      " 10.\n",
      " 0. This verse\n",
      " 5.\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0.\n",
      " 0.\n",
      " 2.\n",
      "\n",
      "\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0.\n",
      " 0. This verse\n",
      " 0.\n",
      " 0. This verse\n",
      " 0.\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0.\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0.\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 10. This\n",
      " 0. This verse\n",
      "# the current counter is 780\n",
      " 8.\n",
      "\n",
      "\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 5.\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0.\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0.\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 2.\n",
      " 2. This verse\n",
      " 10. This\n",
      " 5.\n",
      " 0.\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 10.\n",
      " 10. This\n",
      " 0. This verse\n",
      " 8.\n",
      "\n",
      "\n",
      " 5.\n",
      "\n",
      "\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 8.\n",
      "\n",
      "\n",
      " 8.\n",
      "\n",
      "\n",
      " 10. This\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0.\n",
      " 10. This\n",
      " 5.\n",
      "\n",
      "\n",
      " 0.\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      "# the current counter is 840\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 10.\n",
      "\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0.\n",
      " 0. This verse\n",
      " 0.\n",
      " 0.\n",
      " 0. This verse\n",
      " 0.\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 8.\n",
      "\n",
      "\n",
      " 0.\n",
      " 10. This\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0.\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 10. This\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0.\n",
      " 0. This verse\n",
      "# the current counter is 900\n",
      " 3. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 10. This\n",
      " 10. This\n",
      " 5.\n",
      "\n",
      "\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0.\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0.\n",
      " 0.\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0.\n",
      " 0. This verse\n",
      " 8.\n",
      "\n",
      "\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0.\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0.\n",
      " 0.\n",
      " 0. This verse\n",
      " 0.\n",
      " 10. (\n",
      " 0.\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 5.\n",
      " 8.\n",
      "\n",
      "\n",
      " 5.\n",
      " 0. This verse\n",
      " 3. This verse\n",
      " 8.\n",
      "\n",
      "\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 10. This\n",
      " 0. This verse\n",
      " 0. This verse\n",
      "# the current counter is 960\n",
      " 0. This verse\n",
      " 8.\n",
      "\n",
      "\n",
      " 0.\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 10.\n",
      " 0.\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 10.\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0.\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0.\n",
      " 2.\n",
      "\n",
      "\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0.\n",
      " 0. This verse\n",
      " 0.\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 10. (\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 10.\n",
      " 10. (\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      "# the current counter is 1020\n",
      " 0. This verse\n",
      " 0.\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0.\n",
      " 0. This verse\n",
      " 0.\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0.\n",
      " 10. (\n",
      " 0.\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0.\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0.\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0.\n",
      " 0. This verse\n",
      " 0.\n",
      " 0.\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0.\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0.\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0. This verse\n",
      " 0.\n",
      " 0.\n",
      " 0. This verse\n",
      "all 1071 prompts took: 1201.56 seconds\n"
     ]
    }
   ],
   "source": [
    "# if the column ollama_mistral7B_ratings already exists, then read skip this execution\n",
    "if \"ollama_mistral7B_ratings\" in list(experiment_df.columns):\n",
    "    print(\"ollama_mistral7B_ratings already exists\")\n",
    "else:\n",
    "    # read modelfile into string\n",
    "    model_file = open('./ollama_modelfile/mistral7B_llm').read()\n",
    "    model_name = \"seeker_mistral\"\n",
    "    ollama.create(model=model_name, modelfile=model_file)\n",
    "    \n",
    "    # get the score for each prompt\n",
    "    ollama_mistral7B_responses = []\n",
    "    ollama_mistral7B_response_objects = []\n",
    "    start_timestamp = time.time()\n",
    "    last_start_timestamp = time.time()\n",
    "    counter = 0\n",
    "    for prompt in verses:\n",
    "        counter += 1\n",
    "        if counter%60 == 0:\n",
    "            print(\"# the current counter is\", counter)\n",
    "        response_object = ollama.generate(model=model_name, prompt=prompt)\n",
    "        response = response_object[\"response\"]\n",
    "        print(response)\n",
    "        ollama_mistral7B_responses.append(response)\n",
    "        ollama_mistral7B_response_objects.append(response_object)\n",
    "    \n",
    "    total_time_for_ollama_mistral7B = time.time() - start_timestamp\n",
    "    print(f\"all {len(verses)} prompts took: {total_time_for_ollama_mistral7B:.2f} seconds\")\n",
    "    \n",
    "    ollama_mistral7B_responses_int = [convert_to_int(response.strip().split(\".\")[0]) if response.strip() != \"\" else -3 for response in ollama_mistral7B_responses ]\n",
    "\n",
    "    experiment_df.insert(2,\"ollama_mistral7B_ratings\", ollama_mistral7B_responses_int)\n",
    "    experiment_df.insert(2,\"ollama_mistral7B_response_objects\", ollama_mistral7B_response_objects)\n",
    "    \n",
    "    # save time in times_df\n",
    "    times_df = pd.concat([times_df, pd.DataFrame([{'name': 'ollama_mistral7B', 'sec': total_time_for_ollama_mistral7B}])])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## run with llama2:7b"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "6\n",
      "\n",
      "6\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "6\n",
      "\n",
      "6\n",
      "\n",
      "6\n",
      "\n",
      "6\n",
      "\n",
      "6\n",
      "\n",
      "6\n",
      "\n",
      "6\n",
      "\n",
      "8\n",
      "\n",
      "8\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "8\n",
      "\n",
      "8\n",
      "\n",
      "8\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "8\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "8\n",
      "\n",
      "7\n",
      "\n",
      "6\n",
      "\n",
      "8\n",
      "\n",
      "7\n",
      "\n",
      "8\n",
      "\n",
      "8\n",
      "\n",
      "6\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "8\n",
      "\n",
      "7\n",
      "\n",
      "8\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "6\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "8\n",
      "\n",
      "8\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "8\n",
      "\n",
      "7\n",
      "\n",
      "8\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "8\n",
      "# the current counter is 60\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "6\n",
      "\n",
      "8\n",
      "\n",
      "8\n",
      "\n",
      "8\n",
      "\n",
      "7\n",
      "\n",
      "6\n",
      "\n",
      "7\n",
      "\n",
      "8\n",
      "\n",
      "6\n",
      "\n",
      "8\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "8\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "8\n",
      "\n",
      "6\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "8\n",
      "\n",
      "6\n",
      "\n",
      "7\n",
      "\n",
      "8\n",
      "\n",
      "8\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "8\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "8\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "8\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "8\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "8\n",
      "\n",
      "6\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "8\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "8\n",
      "# the current counter is 120\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "6\n",
      "\n",
      "8\n",
      "\n",
      "5\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "5\n",
      "\n",
      "8\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "8\n",
      "\n",
      "5\n",
      "\n",
      "6\n",
      "\n",
      "8\n",
      "\n",
      "8\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "8\n",
      "\n",
      "8\n",
      "\n",
      "7\n",
      "\n",
      "8\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "8\n",
      "\n",
      "5\n",
      "\n",
      "8\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "8\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "8\n",
      "\n",
      "7\n",
      "\n",
      "8\n",
      "\n",
      "7\n",
      "\n",
      "8\n",
      "\n",
      "8\n",
      "\n",
      "8\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "8\n",
      "\n",
      "7\n",
      "\n",
      "8\n",
      "# the current counter is 180\n",
      "\n",
      "8\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "8\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "8\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "8\n",
      "\n",
      "7\n",
      "\n",
      "8\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "8\n",
      "\n",
      "7\n",
      "\n",
      "8\n",
      "\n",
      "8\n",
      "\n",
      "7\n",
      "\n",
      "8\n",
      "\n",
      "7\n",
      "\n",
      "8\n",
      "\n",
      "8\n",
      "\n",
      "8\n",
      "\n",
      "8\n",
      "\n",
      "6\n",
      "\n",
      "7\n",
      "\n",
      "8\n",
      "\n",
      "8\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "8\n",
      "\n",
      "7\n",
      "\n",
      "6\n",
      "\n",
      "7\n",
      "\n",
      "8\n",
      "\n",
      "8\n",
      "\n",
      "7\n",
      "\n",
      "6\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "6\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "# the current counter is 240\n",
      "\n",
      "7\n",
      "\n",
      "8\n",
      "\n",
      "6\n",
      "\n",
      "8\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "8\n",
      "\n",
      "6\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "8\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "8\n",
      "\n",
      "6\n",
      "\n",
      "8\n",
      "\n",
      "7\n",
      "\n",
      "6\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "8\n",
      "\n",
      "7\n",
      "\n",
      "8\n",
      "\n",
      "8\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "8\n",
      "\n",
      "8\n",
      "\n",
      "7\n",
      "\n",
      "6\n",
      "\n",
      "6\n",
      "\n",
      "8\n",
      "\n",
      "8\n",
      "\n",
      "8\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "8\n",
      "\n",
      "8\n",
      "\n",
      "8\n",
      "\n",
      "5\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "8\n",
      "# the current counter is 300\n",
      "\n",
      "8\n",
      "\n",
      "8\n",
      "\n",
      "8\n",
      "\n",
      "8\n",
      "\n",
      "8\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "8\n",
      "\n",
      "8\n",
      "\n",
      "8\n",
      "\n",
      "7\n",
      "\n",
      "8\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "6\n",
      "\n",
      "8\n",
      "\n",
      "8\n",
      "\n",
      "8\n",
      "\n",
      "8\n",
      "\n",
      "6\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "8\n",
      "\n",
      "8\n",
      "\n",
      "7\n",
      "\n",
      "8\n",
      "\n",
      "7\n",
      "\n",
      "8\n",
      "\n",
      "6\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "8\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "8\n",
      "\n",
      "8\n",
      "\n",
      "8\n",
      "\n",
      "8\n",
      "\n",
      "8\n",
      "\n",
      "8\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "8\n",
      "\n",
      "6\n",
      "\n",
      "7\n",
      "\n",
      "8\n",
      "\n",
      "7\n",
      "\n",
      "8\n",
      "\n",
      "6\n",
      "\n",
      "7\n",
      "\n",
      "8\n",
      "\n",
      "7\n",
      "\n",
      "8\n",
      "\n",
      "6\n",
      "# the current counter is 360\n",
      "\n",
      "8\n",
      "\n",
      "6\n",
      "\n",
      "8\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "8\n",
      "\n",
      "8\n",
      "\n",
      "8\n",
      "\n",
      "7\n",
      "\n",
      "6\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "8\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "8\n",
      "\n",
      "8\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "6\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "5\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "8\n",
      "\n",
      "6\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "6\n",
      "\n",
      "6\n",
      "\n",
      "6\n",
      "\n",
      "6\n",
      "\n",
      "7\n",
      "\n",
      "8\n",
      "\n",
      "6\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "8\n",
      "\n",
      "7\n",
      "\n",
      "6\n",
      "\n",
      "8\n",
      "\n",
      "7\n",
      "# the current counter is 420\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "6\n",
      "\n",
      "6\n",
      "\n",
      "6\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "8\n",
      "\n",
      "7\n",
      "\n",
      "8\n",
      "\n",
      "8\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "8\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "8\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "8\n",
      "\n",
      "8\n",
      "\n",
      "6\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "6\n",
      "\n",
      "6\n",
      "\n",
      "6\n",
      "\n",
      "8\n",
      "\n",
      "6\n",
      "\n",
      "6\n",
      "\n",
      "7\n",
      "\n",
      "6\n",
      "\n",
      "8\n",
      "\n",
      "8\n",
      "\n",
      "6\n",
      "\n",
      "6\n",
      "\n",
      "6\n",
      "\n",
      "6\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "6\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "6\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "6\n",
      "\n",
      "7\n",
      "\n",
      "6\n",
      "# the current counter is 480\n",
      "\n",
      "8\n",
      "\n",
      "8\n",
      "\n",
      "8\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "6\n",
      "\n",
      "7\n",
      "\n",
      "8\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "8\n",
      "\n",
      "7\n",
      "\n",
      "8\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "6\n",
      "\n",
      "7\n",
      "\n",
      "8\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "6\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "8\n",
      "\n",
      "8\n",
      "\n",
      "6\n",
      "\n",
      "8\n",
      "\n",
      "8\n",
      "\n",
      "8\n",
      "\n",
      "6\n",
      "\n",
      "6\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "6\n",
      "\n",
      "6\n",
      "\n",
      "6\n",
      "\n",
      "6\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "6\n",
      "\n",
      "6\n",
      "\n",
      "6\n",
      "\n",
      "6\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "8\n",
      "# the current counter is 540\n",
      "\n",
      "6\n",
      "\n",
      "7\n",
      "\n",
      "6\n",
      "\n",
      "8\n",
      "\n",
      "8\n",
      "\n",
      "7\n",
      "\n",
      "8\n",
      "\n",
      "8\n",
      "\n",
      "7\n",
      "\n",
      "8\n",
      "\n",
      "7\n",
      "\n",
      "8\n",
      "\n",
      "8\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "8\n",
      "\n",
      "7\n",
      "\n",
      "8\n",
      "\n",
      "7\n",
      "\n",
      "8\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "8\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "8\n",
      "\n",
      "8\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "8\n",
      "\n",
      "8\n",
      "\n",
      "7\n",
      "\n",
      "6\n",
      "\n",
      "6\n",
      "\n",
      "7\n",
      "\n",
      "6\n",
      "\n",
      "8\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "8\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "8\n",
      "\n",
      "6\n",
      "\n",
      "8\n",
      "\n",
      "8\n",
      "\n",
      "8\n",
      "\n",
      "7\n",
      "\n",
      "8\n",
      "\n",
      "8\n",
      "\n",
      "8\n",
      "\n",
      "7\n",
      "\n",
      "8\n",
      "# the current counter is 600\n",
      "\n",
      "8\n",
      "\n",
      "8\n",
      "\n",
      "7\n",
      "\n",
      "8\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "8\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "8\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "8\n",
      "\n",
      "7\n",
      "\n",
      "8\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "8\n",
      "\n",
      "8\n",
      "\n",
      "8\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "6\n",
      "\n",
      "6\n",
      "\n",
      "8\n",
      "\n",
      "6\n",
      "\n",
      "8\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "8\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "8\n",
      "\n",
      "6\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "8\n",
      "\n",
      "8\n",
      "\n",
      "7\n",
      "\n",
      "8\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "6\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "6\n",
      "\n",
      "7\n",
      "\n",
      "6\n",
      "\n",
      "6\n",
      "\n",
      "7\n",
      "# the current counter is 660\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "8\n",
      "\n",
      "7\n",
      "\n",
      "8\n",
      "\n",
      "8\n",
      "\n",
      "6\n",
      "\n",
      "8\n",
      "\n",
      "6\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "8\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "8\n",
      "\n",
      "8\n",
      "\n",
      "8\n",
      "\n",
      "8\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "8\n",
      "\n",
      "7\n",
      "\n",
      "8\n",
      "\n",
      "6\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "8\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "6\n",
      "\n",
      "7\n",
      "\n",
      "6\n",
      "\n",
      "6\n",
      "\n",
      "8\n",
      "\n",
      "8\n",
      "\n",
      "6\n",
      "\n",
      "8\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "6\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "6\n",
      "\n",
      "7\n",
      "\n",
      "6\n",
      "\n",
      "7\n",
      "\n",
      "6\n",
      "\n",
      "7\n",
      "# the current counter is 720\n",
      "\n",
      "6\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "8\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "8\n",
      "\n",
      "7\n",
      "\n",
      "6\n",
      "\n",
      "7\n",
      "\n",
      "6\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "6\n",
      "\n",
      "8\n",
      "\n",
      "7\n",
      "\n",
      "6\n",
      "\n",
      "6\n",
      "\n",
      "8\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "6\n",
      "\n",
      "6\n",
      "\n",
      "6\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "6\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "8\n",
      "\n",
      "8\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "8\n",
      "\n",
      "8\n",
      "\n",
      "7\n",
      "\n",
      "8\n",
      "\n",
      "8\n",
      "\n",
      "7\n",
      "\n",
      "8\n",
      "\n",
      "7\n",
      "\n",
      "6\n",
      "\n",
      "7\n",
      "\n",
      "8\n",
      "\n",
      "6\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "6\n",
      "# the current counter is 780\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "8\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "8\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "6\n",
      "\n",
      "8\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "8\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "8\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "8\n",
      "\n",
      "6\n",
      "\n",
      "8\n",
      "\n",
      "7\n",
      "\n",
      "6\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "8\n",
      "\n",
      "8\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "5\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "6\n",
      "# the current counter is 840\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "8\n",
      "\n",
      "8\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "8\n",
      "\n",
      "6\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "6\n",
      "\n",
      "8\n",
      "\n",
      "7\n",
      "\n",
      "8\n",
      "\n",
      "8\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "5\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "6\n",
      "\n",
      "6\n",
      "\n",
      "6\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "6\n",
      "\n",
      "7\n",
      "\n",
      "8\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "6\n",
      "\n",
      "6\n",
      "\n",
      "7\n",
      "\n",
      "8\n",
      "\n",
      "6\n",
      "\n",
      "8\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "8\n",
      "\n",
      "8\n",
      "# the current counter is 900\n",
      "\n",
      "8\n",
      "\n",
      "8\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "6\n",
      "\n",
      "6\n",
      "\n",
      "6\n",
      "\n",
      "7\n",
      "\n",
      "8\n",
      "\n",
      "8\n",
      "\n",
      "7\n",
      "\n",
      "6\n",
      "\n",
      "7\n",
      "\n",
      "6\n",
      "8\n",
      "\n",
      "7\n",
      "\n",
      "6\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "6\n",
      "\n",
      "8\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "8\n",
      "\n",
      "8\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "6\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "8\n",
      "\n",
      "7\n",
      "\n",
      "8\n",
      "\n",
      "8\n",
      "\n",
      "8\n",
      "\n",
      "8\n",
      "\n",
      "8\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "8\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "8\n",
      "\n",
      "8\n",
      "\n",
      "6\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "8\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "# the current counter is 960\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "8\n",
      "\n",
      "8\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "6\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "8\n",
      "\n",
      "7\n",
      "\n",
      "8\n",
      "\n",
      "6\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "6\n",
      "\n",
      "7\n",
      "\n",
      "6\n",
      "\n",
      "6\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "6\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "6\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "8\n",
      "\n",
      "6\n",
      "\n",
      "7\n",
      "\n",
      "6\n",
      "\n",
      "6\n",
      "\n",
      "7\n",
      "\n",
      "6\n",
      "\n",
      "7\n",
      "\n",
      "6\n",
      "\n",
      "6\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "8\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "6\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "6\n",
      "\n",
      "6\n",
      "\n",
      "7\n",
      "\n",
      "6\n",
      "# the current counter is 1020\n",
      "\n",
      "7\n",
      "\n",
      "6\n",
      "\n",
      "7\n",
      "\n",
      "5\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "8\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "6\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "6\n",
      "\n",
      "7\n",
      "\n",
      "6\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "6\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "6\n",
      "\n",
      "6\n",
      "\n",
      "6\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "8\n",
      "\n",
      "8\n",
      "\n",
      "8\n",
      "\n",
      "8\n",
      "\n",
      "7\n",
      "\n",
      "8\n",
      "\n",
      "6\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "8\n",
      "\n",
      "7\n",
      "\n",
      "7\n",
      "\n",
      "6\n",
      "\n",
      "7\n",
      "\n",
      "8\n",
      "\n",
      "8\n",
      "all 1071 prompts took: 1091.68 seconds\n"
     ]
    }
   ],
   "source": [
    "import ollama\n",
    "# read modelfile into string\n",
    "model_file = open('./ollama_modelfile/llama2_7B').read()\n",
    "model_name = \"seeker_llama7B\"\n",
    "ollama.create(model=model_name, modelfile=model_file)\n",
    "\n",
    "# if the column ollama_llama7B_ratings already exists, then read skip this execution\n",
    "if \"ollama_llama7B_ratings\" in list(experiment_df.columns):\n",
    "    print(\"ollama_llama7B_ratings already exists\")\n",
    "else:\n",
    "    # get the score for each prompt\n",
    "    ollama_llama7B_responses = []\n",
    "    ollama_llama7B_response_objects = []\n",
    "    start_timestamp = time.time()\n",
    "    last_start_timestamp = time.time()\n",
    "    counter = 0\n",
    "    for prompt in verses:\n",
    "        counter += 1\n",
    "        if counter%60 == 0:\n",
    "            print(\"# the current counter is\", counter)\n",
    "        response_object = ollama.generate(model=model_name, prompt=prompt)\n",
    "        response = response_object[\"response\"]\n",
    "        print(response)\n",
    "        ollama_llama7B_responses.append(response)\n",
    "        ollama_llama7B_response_objects.append(response_object)\n",
    "\n",
    "    total_time_for_ollama_llama7B = time.time() - start_timestamp\n",
    "    print(f\"all {len(verses)} prompts took: {total_time_for_ollama_llama7B:.2f} seconds\")\n",
    "    \n",
    "    ollama_llama7B_responses_int = [convert_to_int(response.strip().split(\".\")[0]) if response.strip() != \"\" else -3 for response in ollama_llama7B_responses ]\n",
    "\n",
    "    experiment_df.insert(2,\"ollama_llama7B_ratings\", ollama_llama7B_responses_int)\n",
    "    experiment_df.insert(2,\"ollama_llama7B_response_objects\", ollama_llama7B_response_objects)\n",
    "    \n",
    "    # save time in times_df\n",
    "    times_df = pd.concat([times_df, pd.DataFrame([{'name': 'ollama_llama7B', 'sec': total_time_for_ollama_llama7B}])])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "I would rate this\n",
      "\n",
      "5\n",
      "\n",
      "Okay, here\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "I would rate this\n",
      "\n",
      "Okay, here\n",
      "\n",
      "I would rate this\n",
      "\n",
      "I would rate this\n",
      "\n",
      "I would rate this\n",
      "\n",
      "I would rate this\n",
      "\n",
      "I would rate this\n",
      "\n",
      "I would rate the\n",
      "\n",
      "I would rate this\n",
      "\n",
      "I would rate this\n",
      "\n",
      "I would rate this\n",
      "\n",
      "I would rate this\n",
      "\n",
      "5\n",
      "\n",
      "I would rate this\n",
      "\n",
      "I would rate this\n",
      "\n",
      "5\n",
      "\n",
      "Based on the\n",
      "\n",
      "5\n",
      "\n",
      "I would rate this\n",
      "\n",
      "Okay, here\n",
      "\n",
      "Okay, I\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "\n",
      "Okay, here\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "\n",
      "I would rate this\n",
      "\n",
      "Okay, here\n",
      "\n",
      "5\n",
      "\n",
      "Okay, here\n",
      "\n",
      "Okay, here\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "\n",
      "6\n",
      "\n",
      "Based on the\n",
      "\n",
      "5\n",
      "\n",
      "Okay, here\n",
      "\n",
      "5\n",
      "\n",
      "I understand, and\n",
      "\n",
      "Okay, here\n",
      "\n",
      "I would rate this\n",
      "\n",
      "Okay, here\n",
      "\n",
      "5\n",
      "\n",
      "Based on the\n",
      "\n",
      "Okay, here\n",
      "\n",
      "8\n",
      "\n",
      "5\n",
      "\n",
      "Based on the\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "\n",
      "Based on the\n",
      "\n",
      "5\n",
      "\n",
      "Okay, here\n",
      "\n",
      "Okay, I\n",
      "\n",
      "I would rate this\n",
      "# the current counter is 60\n",
      "\n",
      "8\n",
      "\n",
      "5\n",
      "\n",
      "Okay, I\n",
      "\n",
      "5\n",
      "\n",
      "8\n",
      "\n",
      "8\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "\n",
      "Okay, I\n",
      "\n",
      "Okay, I\n",
      "\n",
      "5\n",
      "\n",
      "Okay, I\n",
      "\n",
      "Okay, I\n",
      "\n",
      "8\n",
      "\n",
      "Okay, here\n",
      "\n",
      "Based on the\n",
      "\n",
      "6\n",
      "\n",
      "Okay, here\n",
      "\n",
      "5\n",
      "\n",
      "7\n",
      "\n",
      "Okay, I\n",
      "\n",
      "5\n",
      "\n",
      "Okay, here\n",
      "\n",
      "Okay, here\n",
      "\n",
      "Okay, here\n",
      "\n",
      "6\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "\n",
      "6\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "\n",
      "6\n",
      "\n",
      "8\n",
      "\n",
      "5\n",
      "\n",
      "Okay, here\n",
      "\n",
      "6\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "\n",
      "Okay, here\n",
      "\n",
      "Based on the\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "\n",
      "Okay, I\n",
      "\n",
      "Based on the\n",
      "\n",
      "Okay, I\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "\n",
      "Okay, I\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "# the current counter is 120\n",
      "\n",
      "5\n",
      "\n",
      "Okay, I\n",
      "\n",
      "Based on the\n",
      "\n",
      "Okay, I\n",
      "\n",
      "Okay, I\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "\n",
      "6\n",
      "\n",
      "Okay, I\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "\n",
      "Okay, I\n",
      "\n",
      "7\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "\n",
      "Okay, here\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "\n",
      "Okay, here\n",
      "\n",
      "Okay, here\n",
      "\n",
      "Okay, here\n",
      "\n",
      "5\n",
      "\n",
      "Okay, I\n",
      "\n",
      "6\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "\n",
      "Okay, I\n",
      "\n",
      "5\n",
      "\n",
      "Okay, I\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "\n",
      "Okay, I\n",
      "\n",
      "Okay, I\n",
      "\n",
      "6\n",
      "\n",
      "Okay, I\n",
      "\n",
      "5\n",
      "\n",
      "8\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "\n",
      "Okay, I\n",
      "\n",
      "Okay, I\n",
      "\n",
      "5\n",
      "\n",
      "Okay, here\n",
      "\n",
      "Okay, I\n",
      "\n",
      "5\n",
      "\n",
      "6\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "\n",
      "Okay, here\n",
      "\n",
      "Based on the\n",
      "\n",
      "I would rate this\n",
      "\n",
      "Okay, here\n",
      "\n",
      "Okay, here\n",
      "# the current counter is 180\n",
      "\n",
      "8\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "\n",
      "Okay, I\n",
      "\n",
      "5\n",
      "\n",
      "Okay, I\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "\n",
      "Okay, I\n",
      "\n",
      "Okay, here\n",
      "\n",
      "5\n",
      "\n",
      "Okay, I\n",
      "\n",
      "8\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "\n",
      "Based on the\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "\n",
      "Okay, here\n",
      "\n",
      "5\n",
      "\n",
      "Okay, here\n",
      "\n",
      "I would rate this\n",
      "\n",
      "Okay, I\n",
      "\n",
      "5\n",
      "\n",
      "Okay, here\n",
      "\n",
      "Based on the\n",
      "\n",
      "Okay, I\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "\n",
      "Okay, here\n",
      "\n",
      "Okay, I\n",
      "\n",
      "5\n",
      "\n",
      "Based on the\n",
      "\n",
      "Okay, I\n",
      "\n",
      "5\n",
      "\n",
      "Okay, here\n",
      "\n",
      "Okay, here\n",
      "\n",
      "Okay, I\n",
      "\n",
      "Okay, I\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "\n",
      "Okay, here\n",
      "\n",
      "8\n",
      "\n",
      "I would rate this\n",
      "\n",
      "Okay, here\n",
      "\n",
      "5\n",
      "\n",
      "8\n",
      "\n",
      "Okay, here\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "\n",
      "Okay, here\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "# the current counter is 240\n",
      "\n",
      "5\n",
      "\n",
      "Okay, I\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "\n",
      "Okay, here\n",
      "\n",
      "5\n",
      "\n",
      "Okay, here\n",
      "\n",
      "Okay, I\n",
      "\n",
      "Okay, I\n",
      "\n",
      "Okay, I\n",
      "\n",
      "Okay, I\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "\n",
      "Okay, here\n",
      "\n",
      "5\n",
      "\n",
      "Okay, I\n",
      "\n",
      "5\n",
      "\n",
      "Okay, here\n",
      "\n",
      "Okay, I\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "\n",
      "Okay, here\n",
      "\n",
      "Okay, here\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "\n",
      "Okay, here\n",
      "\n",
      "I would rate this\n",
      "\n",
      "6\n",
      "\n",
      "Okay, here\n",
      "\n",
      "Okay, I\n",
      "\n",
      "5\n",
      "\n",
      "8\n",
      "\n",
      "Okay, I\n",
      "\n",
      "Based on the\n",
      "\n",
      "5\n",
      "\n",
      "Okay, here\n",
      "\n",
      "5\n",
      "\n",
      "Okay, I\n",
      "\n",
      "6\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "\n",
      "Okay, here\n",
      "\n",
      "5\n",
      "\n",
      "8\n",
      "\n",
      "Okay, here\n",
      "\n",
      "6\n",
      "\n",
      "Okay, here\n",
      "\n",
      "5\n",
      "\n",
      "I would rate this\n",
      "\n",
      "Okay, I\n",
      "\n",
      "Based on the\n",
      "\n",
      "6\n",
      "# the current counter is 300\n",
      "\n",
      "5\n",
      "\n",
      "Okay, here\n",
      "\n",
      "Okay, I\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "\n",
      "8\n",
      "\n",
      "Okay, I\n",
      "\n",
      "Okay, here\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "\n",
      "Okay, here\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "\n",
      "Okay, I\n",
      "\n",
      "8\n",
      "\n",
      "I would rate this\n",
      "\n",
      "5\n",
      "\n",
      "Okay, here\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "\n",
      "6\n",
      "\n",
      "Based on the\n",
      "\n",
      "Okay, here\n",
      "\n",
      "5\n",
      "\n",
      "Okay, I\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "\n",
      "6\n",
      "\n",
      "5\n",
      "\n",
      "Okay, I\n",
      "\n",
      "Okay, I\n",
      "\n",
      "5\n",
      "\n",
      "Okay, here\n",
      "\n",
      "5\n",
      "\n",
      "Okay, here\n",
      "\n",
      "5\n",
      "\n",
      "Based on the\n",
      "\n",
      "5\n",
      "\n",
      "Okay, I\n",
      "\n",
      "5\n",
      "\n",
      "Okay, here\n",
      "\n",
      "Okay, I\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "\n",
      "Okay, here\n",
      "\n",
      "Okay, I\n",
      "\n",
      "Okay, I\n",
      "\n",
      "Based on the\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "\n",
      "3\n",
      "\n",
      "Okay, here\n",
      "\n",
      "Okay, I\n",
      "\n",
      "5\n",
      "\n",
      "Okay, here\n",
      "\n",
      "5\n",
      "# the current counter is 360\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "\n",
      "7\n",
      "\n",
      "Okay, here\n",
      "\n",
      "5\n",
      "\n",
      "Okay, here\n",
      "\n",
      "5\n",
      "\n",
      "6\n",
      "\n",
      "5\n",
      "\n",
      "Okay, I\n",
      "\n",
      "Based on the\n",
      "\n",
      "5\n",
      "\n",
      "I would rate this\n",
      "\n",
      "I would rate this\n",
      "\n",
      "Okay, here\n",
      "\n",
      "Okay, here\n",
      "\n",
      "5\n",
      "\n",
      "8\n",
      "\n",
      "Okay, I\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "\n",
      "I would rate this\n",
      "\n",
      "Okay, I\n",
      "\n",
      "Okay, here\n",
      "\n",
      "5\n",
      "\n",
      "Okay, I\n",
      "\n",
      "5\n",
      "\n",
      "Okay, I\n",
      "\n",
      "Based on the\n",
      "\n",
      "5\n",
      "\n",
      "Okay, here\n",
      "\n",
      "Okay, I\n",
      "\n",
      "Okay, I\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "\n",
      "Okay, I\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "\n",
      "Okay, here\n",
      "\n",
      "Okay, I\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "\n",
      "Based on the\n",
      "\n",
      "Based on the\n",
      "\n",
      "5\n",
      "\n",
      "Okay, here\n",
      "\n",
      "Okay, I\n",
      "\n",
      "5\n",
      "\n",
      "Okay, I\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "\n",
      "Okay, I\n",
      "\n",
      "Okay, I\n",
      "# the current counter is 420\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "\n",
      "Based on the\n",
      "\n",
      "Sure! Here\n",
      "\n",
      "Okay, I\n",
      "\n",
      "5\n",
      "\n",
      "Okay, I\n",
      "\n",
      "Okay, I\n",
      "\n",
      "Based on the\n",
      "\n",
      "5\n",
      "\n",
      "Okay, here\n",
      "\n",
      "Okay, here\n",
      "\n",
      "I would rate this\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "\n",
      "8\n",
      "\n",
      "Okay, I\n",
      "\n",
      "5\n",
      "\n",
      "Okay, I\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "\n",
      "Okay, here\n",
      "\n",
      "5\n",
      "\n",
      "Based on the\n",
      "\n",
      "Okay, I\n",
      "\n",
      "Based on the\n",
      "\n",
      "8\n",
      "\n",
      "5\n",
      "\n",
      "Sure! Here\n",
      "\n",
      "Based on the\n",
      "\n",
      "5\n",
      "\n",
      "Okay, here\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "\n",
      "Okay, I\n",
      "\n",
      "Based on the\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "\n",
      "Okay, here\n",
      "\n",
      "5\n",
      "\n",
      "Okay, here\n",
      "\n",
      "Okay, here\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "\n",
      "Okay, here\n",
      "\n",
      "5\n",
      "\n",
      "Okay, here\n",
      "\n",
      "Okay, here\n",
      "# the current counter is 480\n",
      "\n",
      "Okay, here\n",
      "\n",
      "Okay, here\n",
      "\n",
      "Okay, here\n",
      "\n",
      "Okay, here\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "\n",
      "8\n",
      "\n",
      "5\n",
      "\n",
      "Based on the\n",
      "\n",
      "5\n",
      "\n",
      "I would rate this\n",
      "\n",
      "Okay, here\n",
      "\n",
      "5\n",
      "\n",
      "Okay, I\n",
      "\n",
      "Okay, I\n",
      "\n",
      "5\n",
      "\n",
      "I can give you\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "\n",
      "Okay, here\n",
      "\n",
      "Okay, I\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "\n",
      "I would rate this\n",
      "\n",
      "5\n",
      "\n",
      "Okay, here\n",
      "\n",
      "Okay, here\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "\n",
      "Okay, here\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "\n",
      "6\n",
      "\n",
      "Okay, here\n",
      "\n",
      "Okay, here\n",
      "\n",
      "Okay, I\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "\n",
      "Okay, I\n",
      "\n",
      "Okay, here\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "\n",
      "Okay, here\n",
      "\n",
      "5\n",
      "\n",
      "Okay, I\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "\n",
      "Based on the\n",
      "# the current counter is 540\n",
      "\n",
      "5\n",
      "\n",
      "Okay, I\n",
      "\n",
      "Okay, I\n",
      "\n",
      "5\n",
      "\n",
      "6\n",
      "\n",
      "Okay, I\n",
      "\n",
      "Okay, I\n",
      "\n",
      "Okay, I\n",
      "\n",
      "5\n",
      "\n",
      "Okay, here\n",
      "\n",
      "Okay, here\n",
      "\n",
      "Okay, here\n",
      "\n",
      "8\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "\n",
      "Okay, here\n",
      "\n",
      "Okay, here\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "\n",
      "Okay, here\n",
      "\n",
      "Okay, I\n",
      "\n",
      "Sure! Here\n",
      "\n",
      "Okay, here\n",
      "\n",
      "5\n",
      "\n",
      "8\n",
      "\n",
      "Okay, I\n",
      "\n",
      "5\n",
      "\n",
      "Okay, I\n",
      "\n",
      "5\n",
      "\n",
      "Okay, I\n",
      "\n",
      "Okay, I\n",
      "\n",
      "5\n",
      "\n",
      "Okay, I\n",
      "\n",
      "5\n",
      "\n",
      "Based on the\n",
      "\n",
      "Based on the\n",
      "\n",
      "Okay, I\n",
      "\n",
      "Based on the\n",
      "\n",
      "I would rate this\n",
      "\n",
      "5\n",
      "\n",
      "Okay, I\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "\n",
      "Okay, I\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "\n",
      "8\n",
      "\n",
      "5\n",
      "\n",
      "Okay, I\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "\n",
      "Okay, I\n",
      "\n",
      "5\n",
      "# the current counter is 600\n",
      "\n",
      "Based on the\n",
      "\n",
      "5\n",
      "\n",
      "Okay, I\n",
      "\n",
      "5\n",
      "\n",
      "Okay, I\n",
      "\n",
      "Okay, I\n",
      "\n",
      "5\n",
      "\n",
      "Based on the\n",
      "\n",
      "5\n",
      "\n",
      "Okay, here\n",
      "\n",
      "5\n",
      "\n",
      "Okay, I\n",
      "\n",
      "Okay, here\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "\n",
      "Okay, here\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "\n",
      "Okay, here\n",
      "\n",
      "I would rate this\n",
      "\n",
      "Okay, I\n",
      "\n",
      "5\n",
      "\n",
      "Okay, here\n",
      "\n",
      "Okay, I\n",
      "\n",
      "5\n",
      "\n",
      "Okay, here\n",
      "\n",
      "Okay, I\n",
      "\n",
      "Based on the\n",
      "\n",
      "5\n",
      "\n",
      "Okay, here\n",
      "\n",
      "5\n",
      "\n",
      "Okay, here\n",
      "\n",
      "Okay, here\n",
      "\n",
      "Okay, I\n",
      "\n",
      "5\n",
      "\n",
      "Okay, here\n",
      "\n",
      "Okay, here\n",
      "\n",
      "5\n",
      "\n",
      "Okay, I\n",
      "\n",
      "5\n",
      "\n",
      "I would rate this\n",
      "\n",
      "8\n",
      "\n",
      "8\n",
      "\n",
      "Based on the\n",
      "\n",
      "8\n",
      "\n",
      "5\n",
      "\n",
      "Okay, I\n",
      "\n",
      "Based on the\n",
      "\n",
      "5\n",
      "\n",
      "Okay, here\n",
      "\n",
      "5\n",
      "\n",
      "Based on the\n",
      "\n",
      "Okay, here\n",
      "\n",
      "Okay, here\n",
      "\n",
      "Okay, here\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "# the current counter is 660\n",
      "\n",
      "Okay, here\n",
      "\n",
      "Okay, here\n",
      "\n",
      "5\n",
      "\n",
      "Okay, here\n",
      "\n",
      "5\n",
      "\n",
      "I would rate this\n",
      "\n",
      "Okay, I\n",
      "\n",
      "Okay, here\n",
      "\n",
      "5\n",
      "\n",
      "Okay, here\n",
      "\n",
      "Okay, here\n",
      "\n",
      "Based on the\n",
      "\n",
      "5\n",
      "\n",
      "Based on the\n",
      "\n",
      "Okay, I\n",
      "\n",
      "I would rate this\n",
      "\n",
      "6\n",
      "\n",
      "5\n",
      "\n",
      "Okay, I\n",
      "\n",
      "Okay, here\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "\n",
      "6\n",
      "\n",
      "Okay, I\n",
      "\n",
      "Okay, I\n",
      "\n",
      "Okay, I\n",
      "\n",
      "I would rate this\n",
      "\n",
      "Okay, here\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "\n",
      "6\n",
      "\n",
      "Okay, here\n",
      "\n",
      "Okay, I\n",
      "\n",
      "5\n",
      "\n",
      "6\n",
      "\n",
      "Okay, here\n",
      "\n",
      "5\n",
      "\n",
      "Based on the\n",
      "\n",
      "Based on the\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "\n",
      "Okay, here\n",
      "\n",
      "Okay, here\n",
      "\n",
      "Okay, here\n",
      "\n",
      "5\n",
      "\n",
      "Based on the\n",
      "\n",
      "Okay, I\n",
      "\n",
      "Based on the\n",
      "\n",
      "Based on the\n",
      "\n",
      "5\n",
      "\n",
      "Okay, I\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "\n",
      "Based on the\n",
      "\n",
      "5\n",
      "\n",
      "Okay, I\n",
      "\n",
      "Okay, here\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "\n",
      "Okay, here\n",
      "# the current counter is 720\n",
      "\n",
      "Okay, I\n",
      "\n",
      "5\n",
      "\n",
      "I would rate the\n",
      "\n",
      "Okay, here\n",
      "\n",
      "Based on the\n",
      "\n",
      "Okay, I\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "\n",
      "I would rate this\n",
      "\n",
      "Okay, I\n",
      "\n",
      "5\n",
      "\n",
      "Based on the\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "\n",
      "Okay, here\n",
      "\n",
      "Okay, here\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "\n",
      "Sure! Here\n",
      "\n",
      "5\n",
      "\n",
      "Based on the\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "\n",
      "Okay, I\n",
      "\n",
      "Okay, I\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "\n",
      "Okay, here\n",
      "\n",
      "5\n",
      "\n",
      "Okay, I\n",
      "\n",
      "Okay, I\n",
      "\n",
      "Okay, I\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "\n",
      "I would rate this\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "\n",
      "I would rate this\n",
      "\n",
      "8\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "\n",
      "Okay, I\n",
      "\n",
      "5\n",
      "\n",
      "8\n",
      "\n",
      "5\n",
      "\n",
      "Sure, I\n",
      "\n",
      "Okay, I\n",
      "\n",
      "I would rate this\n",
      "\n",
      "Okay, I\n",
      "\n",
      "Okay, here\n",
      "\n",
      "I would rate this\n",
      "\n",
      "5\n",
      "\n",
      "Okay, I\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "\n",
      "Based on the\n",
      "# the current counter is 780\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "\n",
      "I would rate this\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "\n",
      "Okay, here\n",
      "\n",
      "Based on the\n",
      "\n",
      "Based on the\n",
      "\n",
      "Okay, I\n",
      "\n",
      "5\n",
      "\n",
      "Okay, I\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "\n",
      "Based on the\n",
      "\n",
      "5\n",
      "\n",
      "Based on the\n",
      "\n",
      "5\n",
      "\n",
      "Okay, I\n",
      "\n",
      "5\n",
      "\n",
      "Okay, I\n",
      "\n",
      "Okay, here\n",
      "\n",
      "I would rate this\n",
      "\n",
      "I would rate the\n",
      "\n",
      "5\n",
      "\n",
      "Based on the\n",
      "\n",
      "Based on the\n",
      "\n",
      "5\n",
      "\n",
      "Okay, here\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "\n",
      "Okay, here\n",
      "\n",
      "Okay, here\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "\n",
      "Based on the\n",
      "\n",
      "6\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "\n",
      "6\n",
      "\n",
      "5\n",
      "\n",
      "Okay, I\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "\n",
      "Based on the\n",
      "# the current counter is 840\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "\n",
      "Based on the\n",
      "\n",
      "5\n",
      "\n",
      "8\n",
      "\n",
      "Okay, I\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "\n",
      "Okay, I\n",
      "\n",
      "Based on the\n",
      "\n",
      "5\n",
      "\n",
      "I would rate this\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "\n",
      "6\n",
      "\n",
      "Okay, I\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "\n",
      "Sure! Here\n",
      "\n",
      "Sorry, I\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "\n",
      "Okay, I\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "\n",
      "Okay, here\n",
      "\n",
      "5\n",
      "\n",
      "Okay, here\n",
      "\n",
      "Okay, I\n",
      "\n",
      "6\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "\n",
      "Okay, I\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "\n",
      "Okay, I\n",
      "\n",
      "Based on the\n",
      "\n",
      "Okay, here\n",
      "\n",
      "Okay, I\n",
      "\n",
      "Okay, here\n",
      "\n",
      "Okay, I\n",
      "\n",
      "Okay, here\n",
      "\n",
      "Okay, here\n",
      "\n",
      "Sure! Here\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "\n",
      "Okay, I\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "\n",
      "Based on the\n",
      "\n",
      "Okay, I\n",
      "# the current counter is 900\n",
      "\n",
      "8\n",
      "\n",
      "Okay, I\n",
      "\n",
      "6\n",
      "\n",
      "Based on the\n",
      "\n",
      "Okay, here\n",
      "\n",
      "Based on the\n",
      "\n",
      "Okay, I\n",
      "\n",
      "5\n",
      "\n",
      "8\n",
      "\n",
      "Okay, here\n",
      "\n",
      "5\n",
      "\n",
      "I would rate this\n",
      "\n",
      "Okay, here\n",
      "\n",
      "I would rate this\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "\n",
      "Okay, I\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "\n",
      "Okay, I\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "\n",
      "Okay, here\n",
      "\n",
      "I would rate this\n",
      "\n",
      "Okay, here\n",
      "\n",
      "5\n",
      "\n",
      "Okay, I\n",
      "\n",
      "Okay, here\n",
      "\n",
      "6\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "\n",
      "Okay, I\n",
      "\n",
      "I would rate this\n",
      "\n",
      "Okay, here\n",
      "\n",
      "Okay, here\n",
      "\n",
      "Sure! Here\n",
      "\n",
      "5\n",
      "\n",
      "Okay, here\n",
      "\n",
      "5\n",
      "\n",
      "Okay, here\n",
      "\n",
      "5\n",
      "\n",
      "Okay, here\n",
      "\n",
      "Okay, here\n",
      "\n",
      "5\n",
      "\n",
      "Okay, here\n",
      "\n",
      "5\n",
      "\n",
      "Okay, I\n",
      "\n",
      "Okay, here\n",
      "\n",
      "Okay, I\n",
      "\n",
      "5\n",
      "\n",
      "Okay, here\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "\n",
      "Okay, here\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "\n",
      "Okay, I\n",
      "\n",
      "5\n",
      "# the current counter is 960\n",
      "\n",
      "Okay, here\n",
      "\n",
      "I would rate this\n",
      "\n",
      "Okay, I\n",
      "\n",
      "5\n",
      "\n",
      "I would rate the\n",
      "\n",
      "Okay, here\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "\n",
      "Based on the\n",
      "\n",
      "5\n",
      "\n",
      "Okay, I\n",
      "\n",
      "Okay, here\n",
      "\n",
      "Okay, here\n",
      "\n",
      "Okay, here\n",
      "\n",
      "Okay, here\n",
      "\n",
      "I would rate this\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "\n",
      "Okay, here\n",
      "\n",
      "5\n",
      "\n",
      "Based on the\n",
      "\n",
      "5\n",
      "\n",
      "I would rate this\n",
      "\n",
      "5\n",
      "\n",
      "Okay, here\n",
      "\n",
      "Okay, here\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "\n",
      "Okay, I\n",
      "\n",
      "5\n",
      "\n",
      "Okay, here\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "\n",
      "Based on the\n",
      "\n",
      "Okay, here\n",
      "\n",
      "Okay, here\n",
      "\n",
      "5\n",
      "\n",
      "Based on the\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "\n",
      "Based on the\n",
      "\n",
      "5\n",
      "\n",
      "Okay, here\n",
      "\n",
      "5\n",
      "\n",
      "Okay, I\n",
      "\n",
      "Okay, here\n",
      "\n",
      "Okay, here\n",
      "\n",
      "Okay, here\n",
      "\n",
      "I would rate this\n",
      "\n",
      "5\n",
      "\n",
      "I would rate this\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "\n",
      "3\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "# the current counter is 1020\n",
      "\n",
      "I would rate this\n",
      "\n",
      "3\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "\n",
      "3\n",
      "\n",
      "Okay, here\n",
      "\n",
      "I would rate this\n",
      "\n",
      "Okay, I\n",
      "\n",
      "8\n",
      "\n",
      "5\n",
      "\n",
      "Based on the\n",
      "\n",
      "Okay, I\n",
      "\n",
      "Okay, I\n",
      "\n",
      "5\n",
      "\n",
      "Okay, here\n",
      "\n",
      "5\n",
      "\n",
      "6\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "\n",
      "Okay, here\n",
      "\n",
      "5\n",
      "\n",
      "Based on the\n",
      "\n",
      "Okay, I\n",
      "\n",
      "Based on the\n",
      "\n",
      "Okay, here\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "\n",
      "Okay, here\n",
      "\n",
      "5\n",
      "\n",
      "Okay, here\n",
      "\n",
      "5\n",
      "\n",
      "Okay, here\n",
      "\n",
      "Okay, here\n",
      "\n",
      "5\n",
      "\n",
      "I would rate this\n",
      "\n",
      "Okay, I\n",
      "\n",
      "Based on the\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "\n",
      "5\n",
      "\n",
      "Okay, I\n",
      "\n",
      "5\n",
      "\n",
      "8\n",
      "all 1071 prompts took: 1091.68 seconds\n"
     ]
    }
   ],
   "source": [
    "# read modelfile into string\n",
    "model_file = open('./ollama_modelfile/llama2_13B').read()\n",
    "model_name = \"seeker_llama13B\"\n",
    "ollama.create(model=model_name, modelfile=model_file)\n",
    "\n",
    "# if the column ollama_llama13B_ratings already exists, then read skip this execution\n",
    "if \"ollama_llama13B_ratings\" in list(experiment_df.columns):\n",
    "    print(\"ollama_llama13B_ratings already exists\")\n",
    "else:\n",
    "    # get the score for each prompt\n",
    "    ollama_llama13B_responses = []\n",
    "    ollama_llama13B_response_objects = []\n",
    "    start_timestamp = time.time()\n",
    "    last_start_timestamp = time.time()\n",
    "    counter = 0\n",
    "    for prompt in verses:\n",
    "        counter += 1\n",
    "        if counter%60 == 0:\n",
    "            print(\"# the current counter is\", counter)\n",
    "        response_object = ollama.generate(model=model_name, prompt=prompt)\n",
    "        response = response_object[\"response\"]\n",
    "        print(response)\n",
    "        ollama_llama13B_responses.append(response)\n",
    "        ollama_llama13B_response_objects.append(response_object)\n",
    "\n",
    "    total_time_for_ollama_llama13B = time.time() - start_timestamp\n",
    "    print(f\"all {len(verses)} prompts took: {total_time_for_ollama_llama7B:.2f} seconds\")\n",
    "    \n",
    "    ollama_llama13B_responses_int = [convert_to_int(response.strip().split(\".\")[0]) if response.strip() != \"\" else -3 for response in ollama_llama13B_responses ]\n",
    "\n",
    "    experiment_df.insert(2,\"ollama_llama13B_ratings\", ollama_llama13B_responses_int)\n",
    "    experiment_df.insert(2,\"ollama_llama13B_response_objects\", ollama_llama13B_response_objects)\n",
    "    \n",
    "    # save time in times_df\n",
    "    times_df = pd.concat([times_df, pd.DataFrame([{'name': 'ollama_llama13B', 'sec': total_time_for_ollama_llama13B}])])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## run with mistral.ai api\n",
    "\n",
    "### Chat Completions API\n",
    "| Model |\t    Input               |\tOutput       |\n",
    "|---------------|-------------------|-----------------|\n",
    "|mistral-tiny\t|0.14€ / 1M tokens\t|0.42€ / 1M tokens|\n",
    "|mistral-small\t|0.6€ / 1M tokens\t|1.8€ / 1M tokens|\n",
    "|mistral-medium\t|2.5€ / 1M tokens\t|7.5€ / 1M tokens|"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {},
   "outputs": [],
   "source": [
    "# set the openai Api key in the enviroment with\n",
    "# setx MISTRAL_API_KEY \"your-api-key-here\"\n",
    "\n",
    "# or in python with\n",
    "# os.environ[\"MISTRAL_API_KEY\"] = \"your-api-key-here\""
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'0'"
      ]
     },
     "execution_count": 37,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from mistralai.client import MistralClient\n",
    "from mistralai.models.chat_completion import ChatMessage\n",
    "\n",
    "api_key = os.environ[\"MISTRAL_API_KEY\"]\n",
    "model = \"mistral-tiny\"\n",
    "\n",
    "client = MistralClient(api_key=api_key)\n",
    "\n",
    "# No streaming\n",
    "def get_mistral_response(model, prompt):\n",
    "    chat_response_object = client.chat(\n",
    "        model=model,\n",
    "        messages=[ChatMessage(role=\"system\", content=sys_prompt), ChatMessage(role=\"user\", content=prompt)],\n",
    "        temperature=0,\n",
    "        max_tokens=2\n",
    "    )\n",
    "    return chat_response_object\n",
    "get_mistral_response(model, verses[10]).choices[0].message.content"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0\n",
      "0\n",
      "0\n",
      "0\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "3\n",
      "0\n",
      "5\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "5\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "1\n",
      "5\n",
      "0\n",
      "1\n",
      "2\n",
      "1\n",
      "0\n",
      "0\n",
      "0\n",
      "5\n",
      "0\n",
      "8\n",
      "0\n",
      "5\n",
      "1\n",
      "0\n",
      "0\n",
      "8\n",
      "0\n",
      "0\n",
      "8\n",
      "0\n",
      "# the current counter is 60\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "1\n",
      "1\n",
      "3\n",
      "5\n",
      "8\n",
      "2\n",
      "0\n",
      "8\n",
      "8\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "1\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "8\n",
      "0\n",
      "0\n",
      "0\n",
      "8\n",
      "1\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "5\n",
      "5\n",
      "0\n",
      "0\n",
      "0\n",
      "5\n",
      "1\n",
      "# the current counter is 120\n",
      "8\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "5\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "1\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "8\n",
      "2\n",
      "5\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "8\n",
      "8\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "5\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "8\n",
      "0\n",
      "1\n",
      "5\n",
      "# the current counter is 180\n",
      "5\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "1\n",
      "8\n",
      "0\n",
      "8\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "1\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "5\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "1\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "1\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "8\n",
      "0\n",
      "0\n",
      "8\n",
      "8\n",
      "0\n",
      "8\n",
      "8\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "# the current counter is 240\n",
      "5\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "5\n",
      "5\n",
      "2\n",
      "0\n",
      "5\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "5\n",
      "2\n",
      "0\n",
      "5\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "8\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "3\n",
      "0\n",
      "5\n",
      "8\n",
      "5\n",
      "0\n",
      "0\n",
      "1\n",
      "1\n",
      "0\n",
      "0\n",
      "5\n",
      "0\n",
      "# the current counter is 300\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "1\n",
      "1\n",
      "0\n",
      "0\n",
      "8\n",
      "5\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "5\n",
      "2\n",
      "5\n",
      "2\n",
      "0\n",
      "8\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "8\n",
      "8\n",
      "3\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "1\n",
      "0\n",
      "0\n",
      "5\n",
      "0\n",
      "5\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "8\n",
      "0\n",
      "0\n",
      "1\n",
      "# the current counter is 360\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "8\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "8\n",
      "0\n",
      "0\n",
      "0\n",
      "8\n",
      "2\n",
      "8\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "1\n",
      "8\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "8\n",
      "0\n",
      "0\n",
      "5\n",
      "8\n",
      "1\n",
      "0\n",
      "0\n",
      "# the current counter is 420\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "5\n",
      "1\n",
      "0\n",
      "0\n",
      "8\n",
      "8\n",
      "0\n",
      "0\n",
      "0\n",
      "1\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "5\n",
      "0\n",
      "0\n",
      "3\n",
      "0\n",
      "0\n",
      "0\n",
      "1\n",
      "0\n",
      "0\n",
      "5\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "3\n",
      "0\n",
      "8\n",
      "0\n",
      "5\n",
      "# the current counter is 480\n",
      "2\n",
      "5\n",
      "0\n",
      "5\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "2\n",
      "0\n",
      "0\n",
      "0\n",
      "3\n",
      "0\n",
      "5\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "8\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "2\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "2\n",
      "0\n",
      "0\n",
      "5\n",
      "0\n",
      "2\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "# the current counter is 540\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "1\n",
      "0\n",
      "5\n",
      "1\n",
      "8\n",
      "1\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "5\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "8\n",
      "0\n",
      "0\n",
      "1\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "8\n",
      "5\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "1\n",
      "8\n",
      "1\n",
      "1\n",
      "0\n",
      "0\n",
      "1\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "# the current counter is 600\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "8\n",
      "0\n",
      "0\n",
      "8\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "1\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "3\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "5\n",
      "8\n",
      "8\n",
      "5\n",
      "1\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "1\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "# the current counter is 660\n",
      "8\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "1\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "2\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "I'\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "8\n",
      "8\n",
      "0\n",
      "3\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "5\n",
      "0\n",
      "0\n",
      "2\n",
      "0\n",
      "0\n",
      "8\n",
      "0\n",
      "0\n",
      "# the current counter is 720\n",
      "1\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "5\n",
      "0\n",
      "8\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "5\n",
      "0\n",
      "1\n",
      "0\n",
      "1\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "2\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "I the\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "1\n",
      "8\n",
      "# the current counter is 780\n",
      "8\n",
      "2\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "5\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "8\n",
      "5\n",
      "1\n",
      "5\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "8\n",
      "1\n",
      "1\n",
      "0\n",
      "8\n",
      "1\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "8\n",
      "8\n",
      "1\n",
      "5\n",
      "0\n",
      "5\n",
      "0\n",
      "0\n",
      "0\n",
      "3\n",
      "0\n",
      "8\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "8\n",
      "0\n",
      "0\n",
      "# the current counter is 840\n",
      "0\n",
      "0\n",
      "1\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "5\n",
      "8\n",
      "0\n",
      "8\n",
      "0\n",
      "0\n",
      "5\n",
      "0\n",
      "0\n",
      "5\n",
      "0\n",
      "2\n",
      "8\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "8\n",
      "0\n",
      "0\n",
      "2\n",
      "0\n",
      "0\n",
      "8\n",
      "0\n",
      "5\n",
      "0\n",
      "0\n",
      "0\n",
      "# the current counter is 900\n",
      "5\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "1\n",
      "1\n",
      "0\n",
      "1\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "1\n",
      "0\n",
      "0\n",
      "0\n",
      "3\n",
      "0\n",
      "0\n",
      "0\n",
      "1\n",
      "0\n",
      "5\n",
      "3\n",
      "8\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "8\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "8\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "8\n",
      "1\n",
      "5\n",
      "0\n",
      "3\n",
      "8\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "1\n",
      "0\n",
      "0\n",
      "# the current counter is 960\n",
      "0\n",
      "8\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "1\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "1\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "3\n",
      "1\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "5\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "2\n",
      "1\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "8\n",
      "1\n",
      "0\n",
      "0\n",
      "1\n",
      "1\n",
      "0\n",
      "8\n",
      "0\n",
      "8\n",
      "# the current counter is 1020\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "1\n",
      "0\n",
      "5\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "5\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "all 1071 prompts took: 3814.5973172187805 seconds\n"
     ]
    }
   ],
   "source": [
    "# if the column mistral_tiny_ratings already exists, then read skip this execution\n",
    "if \"mistral_tiny_ratings\" in list(experiment_df.columns):\n",
    "    print(\"mistral_tiny_ratings already exists\")\n",
    "else:\n",
    "    model = \"mistral-tiny\"\n",
    "    # get the score for each prompt\n",
    "    mistral_tiny_responses = []\n",
    "    mistral_tiny_response_objects = []\n",
    "    start_timestamp = time.time()\n",
    "    last_start_timestamp = time.time()\n",
    "    counter = 0\n",
    "    for prompt in verses:\n",
    "        counter += 1\n",
    "        if counter%60 == 0:\n",
    "            print(\"# the current counter is\", counter)\n",
    "        response = get_mistral_response(model, prompt)\n",
    "        print(response.choices[0].message.content)\n",
    "        mistral_tiny_responses.append(response.choices[0].message.content)\n",
    "        mistral_tiny_response_objects.append(response)\n",
    "    \n",
    "    total_time_for_mistral_tiny = time.time() - start_timestamp\n",
    "    print(\"all\", len(verses), \"prompts took:\", total_time_for_mistral_tiny, \"seconds\")\n",
    "    \n",
    "    mistral_tiny_responses_int = [convert_to_int(response.strip()) if response.strip() != \"\" else -3 for response in mistral_tiny_responses ]\n",
    "\n",
    "    experiment_df.insert(2,\"mistral_tiny_ratings\", mistral_tiny_responses_int)\n",
    "    experiment_df.insert(2,\"mistral_tiny_response_objects\", mistral_tiny_response_objects)\n",
    "    \n",
    "    # save time in times_df\n",
    "    times_df = pd.concat([times_df, pd.DataFrame([{'name': 'mistral_tiny', 'sec': total_time_for_mistral_tiny}])])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "I would\n",
      "1\n",
      "0\n",
      "0\n",
      "1\n",
      "1\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "1\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "I would\n",
      "0\n",
      "1\n",
      "0\n",
      "0\n",
      "\"Be\n",
      "0\n",
      "1\n",
      "0\n",
      "1\n",
      "0\n",
      "0\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "2\n",
      "1\n",
      "0\n",
      "0\n",
      "2\n",
      "8\n",
      "8\n",
      "1\n",
      "0\n",
      "1\n",
      "2\n",
      "0\n",
      "2\n",
      "1\n",
      "8\n",
      "I would\n",
      "2\n",
      "1\n",
      "2\n",
      "9\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "# the current counter is 60\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "0\n",
      "9\n",
      "8\n",
      "0\n",
      "2\n",
      "1\n",
      "2\n",
      "2\n",
      "0\n",
      "0\n",
      "1\n",
      "0\n",
      "2\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "2\n",
      "1\n",
      "2\n",
      "1\n",
      "0\n",
      "1\n",
      "1\n",
      "0\n",
      "1\n",
      "2\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "8\n",
      "9\n",
      "8\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "0\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "0\n",
      "1\n",
      "2\n",
      "0\n",
      "1\n",
      "0\n",
      "0\n",
      "2\n",
      "# the current counter is 120\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "I would\n",
      "1\n",
      "1\n",
      "9\n",
      "8\n",
      "8\n",
      "1\n",
      "1\n",
      "2\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "0\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "0\n",
      "1\n",
      "1\n",
      "2\n",
      "1\n",
      "2\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "2\n",
      "1\n",
      "2\n",
      "2\n",
      "1\n",
      "1\n",
      "2\n",
      "1\n",
      "2\n",
      "0\n",
      "0\n",
      "1\n",
      "1\n",
      "2\n",
      "1\n",
      "1\n",
      "2\n",
      "# the current counter is 180\n",
      "8\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "8\n",
      "0\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "0\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "0\n",
      "1\n",
      "0\n",
      "1\n",
      "1\n",
      "0\n",
      "0\n",
      "1\n",
      "9\n",
      "0\n",
      "8\n",
      "1\n",
      "2\n",
      "1\n",
      "1\n",
      "1\n",
      "0\n",
      "2\n",
      "0\n",
      "0\n",
      "1\n",
      "1\n",
      "0\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "# the current counter is 240\n",
      "1\n",
      "1\n",
      "0\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "2\n",
      "1\n",
      "1\n",
      "7\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "0\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "0\n",
      "0\n",
      "8\n",
      "1\n",
      "2\n",
      "1\n",
      "1\n",
      "1\n",
      "0\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "8\n",
      "1\n",
      "1\n",
      "1\n",
      "2\n",
      "0\n",
      "9\n",
      "9\n",
      "2\n",
      "1\n",
      "1\n",
      "0\n",
      "8\n",
      "2\n",
      "1\n",
      "1\n",
      "1\n",
      "# the current counter is 300\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "9\n",
      "8\n",
      "0\n",
      "2\n",
      "9\n",
      "9\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "0\n",
      "1\n",
      "1\n",
      "8\n",
      "1\n",
      "2\n",
      "1\n",
      "2\n",
      "1\n",
      "0\n",
      "0\n",
      "1\n",
      "2\n",
      "2\n",
      "1\n",
      "1\n",
      "1\n",
      "0\n",
      "1\n",
      "0\n",
      "1\n",
      "1\n",
      "1\n",
      "0\n",
      "0\n",
      "1\n",
      "2\n",
      "0\n",
      "I would\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "0\n",
      "# the current counter is 360\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "0\n",
      "1\n",
      "1\n",
      "0\n",
      "1\n",
      "0\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "0\n",
      "0\n",
      "1\n",
      "1\n",
      "I would\n",
      "1\n",
      "1\n",
      "0\n",
      "2\n",
      "1\n",
      "1\n",
      "1\n",
      "8\n",
      "2\n",
      "0\n",
      "0\n",
      "0\n",
      "2\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "0\n",
      "2\n",
      "2\n",
      "7\n",
      "0\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "2\n",
      "8\n",
      "2\n",
      "1\n",
      "1\n",
      "# the current counter is 420\n",
      "2\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "I would\n",
      "1\n",
      "I would\n",
      "1\n",
      "0\n",
      "1\n",
      "1\n",
      "2\n",
      "8\n",
      "1\n",
      "1\n",
      "0\n",
      "0\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "0\n",
      "0\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "0\n",
      "1\n",
      "1\n",
      "0\n",
      "1\n",
      "0\n",
      "1\n",
      "0\n",
      "0\n",
      "0\n",
      "2\n",
      "1\n",
      "1\n",
      "1\n",
      "0\n",
      "1\n",
      "0\n",
      "0\n",
      "0\n",
      "2\n",
      "2\n",
      "0\n",
      "0\n",
      "0\n",
      "# the current counter is 480\n",
      "0\n",
      "1\n",
      "8\n",
      "2\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "2\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "2\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "0\n",
      "1\n",
      "9\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "0\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "0\n",
      "0\n",
      "0\n",
      "1\n",
      "1\n",
      "0\n",
      "0\n",
      "0\n",
      "1\n",
      "0\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "0\n",
      "1\n",
      "I would\n",
      "0\n",
      "2\n",
      "# the current counter is 540\n",
      "0\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "I answer\n",
      "1\n",
      "0\n",
      "2\n",
      "1\n",
      "0\n",
      "9\n",
      "9\n",
      "2\n",
      "1\n",
      "0\n",
      "2\n",
      "0\n",
      "1\n",
      "1\n",
      "0\n",
      "0\n",
      "1\n",
      "1\n",
      "0\n",
      "1\n",
      "1\n",
      "1\n",
      "0\n",
      "1\n",
      "0\n",
      "1\n",
      "1\n",
      "0\n",
      "1\n",
      "1\n",
      "2\n",
      "1\n",
      "0\n",
      "0\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "0\n",
      "2\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "9\n",
      "2\n",
      "1\n",
      "1\n",
      "9\n",
      "2\n",
      "1\n",
      "1\n",
      "1\n",
      "# the current counter is 600\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "0\n",
      "1\n",
      "I would\n",
      "1\n",
      "0\n",
      "1\n",
      "1\n",
      "0\n",
      "1\n",
      "0\n",
      "1\n",
      "1\n",
      "1\n",
      "0\n",
      "1\n",
      "1\n",
      "1\n",
      "0\n",
      "0\n",
      "1\n",
      "1\n",
      "0\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "2\n",
      "0\n",
      "0\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "9\n",
      "1\n",
      "8\n",
      "1\n",
      "1\n",
      "1\n",
      "5\n",
      "1\n",
      "8\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "# the current counter is 660\n",
      "2\n",
      "1\n",
      "1\n",
      "1\n",
      "2\n",
      "1\n",
      "0\n",
      "0\n",
      "1\n",
      "1\n",
      "2\n",
      "1\n",
      "0\n",
      "1\n",
      "2\n",
      "2\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "I would\n",
      "1\n",
      "1\n",
      "1\n",
      "0\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "2\n",
      "1\n",
      "1\n",
      "0\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "7\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "0\n",
      "1\n",
      "1\n",
      "# the current counter is 720\n",
      "0\n",
      "0\n",
      "I would\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "0\n",
      "2\n",
      "1\n",
      "1\n",
      "0\n",
      "1\n",
      "0\n",
      "0\n",
      "0\n",
      "1\n",
      "2\n",
      "1\n",
      "1\n",
      "1\n",
      "0\n",
      "1\n",
      "0\n",
      "1\n",
      "0\n",
      "0\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "0\n",
      "I would\n",
      "0\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "0\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "2\n",
      "0\n",
      "# the current counter is 780\n",
      "1\n",
      "0\n",
      "I would\n",
      "1\n",
      "1\n",
      "1\n",
      "2\n",
      "0\n",
      "0\n",
      "0\n",
      "1\n",
      "1\n",
      "0\n",
      "1\n",
      "1\n",
      "1\n",
      "0\n",
      "0\n",
      "1\n",
      "1\n",
      "1\n",
      "0\n",
      "0\n",
      "0\n",
      "1\n",
      "1\n",
      "1\n",
      "0\n",
      "2\n",
      "0\n",
      "I would\n",
      "0\n",
      "0\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "0\n",
      "1\n",
      "0\n",
      "7\n",
      "2\n",
      "0\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "2\n",
      "2\n",
      "8\n",
      "0\n",
      "1\n",
      "0\n",
      "1\n",
      "1\n",
      "0\n",
      "1\n",
      "1\n",
      "# the current counter is 840\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "1\n",
      "1\n",
      "I would\n",
      "1\n",
      "1\n",
      "0\n",
      "1\n",
      "1\n",
      "0\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "2\n",
      "1\n",
      "1\n",
      "0\n",
      "7\n",
      "1\n",
      "1\n",
      "1\n",
      "0\n",
      "2\n",
      "1\n",
      "1\n",
      "2\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "2\n",
      "2\n",
      "0\n",
      "1\n",
      "1\n",
      "1\n",
      "2\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "# the current counter is 900\n",
      "1\n",
      "2\n",
      "1\n",
      "2\n",
      "1\n",
      "0\n",
      "0\n",
      "8\n",
      "2\n",
      "1\n",
      "0\n",
      "\"You\n",
      "0\n",
      "0\n",
      "0\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "2\n",
      "1\n",
      "0\n",
      "1\n",
      "0\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "0\n",
      "0\n",
      "0\n",
      "1\n",
      "1\n",
      "1\n",
      "0\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "0\n",
      "1\n",
      "1\n",
      "2\n",
      "1\n",
      "1\n",
      "0\n",
      "1\n",
      "2\n",
      "0\n",
      "1\n",
      "0\n",
      "1\n",
      "0\n",
      "0\n",
      "1\n",
      "# the current counter is 960\n",
      "1\n",
      "0\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "0\n",
      "2\n",
      "0\n",
      "1\n",
      "1\n",
      "1\n",
      "0\n",
      "0\n",
      "1\n",
      "1\n",
      "0\n",
      "1\n",
      "1\n",
      "0\n",
      "1\n",
      "0\n",
      "1\n",
      "0\n",
      "2\n",
      "0\n",
      "0\n",
      "0\n",
      "1\n",
      "0\n",
      "1\n",
      "1\n",
      "0\n",
      "1\n",
      "1\n",
      "1\n",
      "0\n",
      "1\n",
      "1\n",
      "1\n",
      "0\n",
      "0\n",
      "1\n",
      "1\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "1\n",
      "0\n",
      "0\n",
      "0\n",
      "1\n",
      "0\n",
      "9\n",
      "0\n",
      "2\n",
      "1\n",
      "1\n",
      "# the current counter is 1020\n",
      "0\n",
      "1\n",
      "0\n",
      "0\n",
      "1\n",
      "0\n",
      "1\n",
      "0\n",
      "1\n",
      "1\n",
      "0\n",
      "0\n",
      "1\n",
      "1\n",
      "1\n",
      "0\n",
      "1\n",
      "I would\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "0\n",
      "0\n",
      "1\n",
      "1\n",
      "1\n",
      "0\n",
      "0\n",
      "1\n",
      "0\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "all 1071 prompts took: 908.788969039917 seconds\n"
     ]
    }
   ],
   "source": [
    "# if the column mistral_small_ratings already exists, then read skip this execution\n",
    "if \"mistral_small_ratings\" in list(experiment_df.columns):\n",
    "    print(\"mistral_small_ratings already exists\")\n",
    "else:\n",
    "    model = \"mistral-small\"\n",
    "    # get the score for each prompt\n",
    "    mistral_small_responses = []\n",
    "    mistral_small_response_objects = []\n",
    "    start_timestamp = time.time()\n",
    "    last_start_timestamp = time.time()\n",
    "    counter = 0\n",
    "    for prompt in verses:\n",
    "        counter += 1\n",
    "        if counter%60 == 0:\n",
    "            print(\"# the current counter is\", counter)\n",
    "        response = get_mistral_response(model, prompt)\n",
    "        print(response.choices[0].message.content)\n",
    "        mistral_small_responses.append(response.choices[0].message.content)\n",
    "        mistral_small_response_objects.append(response)\n",
    "    \n",
    "    total_time_for_mistral_small = time.time() - start_timestamp\n",
    "    print(\"all\", len(verses), \"prompts took:\", total_time_for_mistral_small, \"seconds\")\n",
    "    \n",
    "    mistral_small_responses_int = [convert_to_int(response.strip()) if response.strip() != \"\" else -3 for response in mistral_small_responses ]\n",
    "\n",
    "    experiment_df.insert(2,\"mistral_small_ratings\", mistral_small_responses_int)\n",
    "    experiment_df.insert(2,\"mistral_small_response_objects\", mistral_small_response_objects)\n",
    "    \n",
    "    # save time in times_df\n",
    "    times_df = pd.concat([times_df, pd.DataFrame([{'name': 'mistral_small', 'sec': total_time_for_mistral_small}])])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "1\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "1\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "0\n",
      "1\n",
      "1\n",
      "0\n",
      "1\n",
      "0\n",
      "2\n",
      "0\n",
      "0\n",
      "1\n",
      "0\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "7\n",
      "1\n",
      "1\n",
      "1\n",
      "0\n",
      "7\n",
      "0\n",
      "1\n",
      "0\n",
      "3\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "7\n",
      "0\n",
      "1\n",
      "3\n",
      "2\n",
      "1\n",
      "2\n",
      "2\n",
      "# the current counter is 60\n",
      "3\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "0\n",
      "7\n",
      "8\n",
      "1\n",
      "1\n",
      "1\n",
      "7\n",
      "2\n",
      "2\n",
      "3\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "0\n",
      "0\n",
      "1\n",
      "2\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "2\n",
      "0\n",
      "1\n",
      "0\n",
      "1\n",
      "3\n",
      "1\n",
      "3\n",
      "1\n",
      "1\n",
      "1\n",
      "8\n",
      "8\n",
      "5\n",
      "3\n",
      "1\n",
      "2\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "3\n",
      "1\n",
      "1\n",
      "1\n",
      "2\n",
      "3\n",
      "3\n",
      "0\n",
      "1\n",
      "8\n",
      "# the current counter is 120\n",
      "8\n",
      "1\n",
      "1\n",
      "1\n",
      "3\n",
      "0\n",
      "1\n",
      "3\n",
      "1\n",
      "7\n",
      "3\n",
      "8\n",
      "2\n",
      "1\n",
      "8\n",
      "3\n",
      "1\n",
      "2\n",
      "2\n",
      "1\n",
      "2\n",
      "1\n",
      "2\n",
      "1\n",
      "2\n",
      "1\n",
      "1\n",
      "0\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "7\n",
      "1\n",
      "2\n",
      "2\n",
      "2\n",
      "1\n",
      "1\n",
      "2\n",
      "3\n",
      "3\n",
      "1\n",
      "1\n",
      "1\n",
      "0\n",
      "2\n",
      "3\n",
      "1\n",
      "2\n",
      "1\n",
      "1\n",
      "3\n",
      "3\n",
      "2\n",
      "3\n",
      "2\n",
      "5\n",
      "# the current counter is 180\n",
      "5\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "7\n",
      "2\n",
      "1\n",
      "2\n",
      "1\n",
      "1\n",
      "2\n",
      "1\n",
      "2\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "2\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "0\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "7\n",
      "1\n",
      "0\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "8\n",
      "1\n",
      "5\n",
      "1\n",
      "5\n",
      "1\n",
      "3\n",
      "1\n",
      "3\n",
      "3\n",
      "0\n",
      "1\n",
      "3\n",
      "1\n",
      "3\n",
      "0\n",
      "1\n",
      "1\n",
      "1\n",
      "# the current counter is 240\n",
      "2\n",
      "3\n",
      "0\n",
      "1\n",
      "1\n",
      "1\n",
      "3\n",
      "3\n",
      "3\n",
      "5\n",
      "3\n",
      "2\n",
      "5\n",
      "1\n",
      "0\n",
      "1\n",
      "2\n",
      "1\n",
      "3\n",
      "1\n",
      "0\n",
      "0\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "2\n",
      "3\n",
      "3\n",
      "3\n",
      "0\n",
      "0\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "2\n",
      "3\n",
      "5\n",
      "1\n",
      "1\n",
      "1\n",
      "2\n",
      "2\n",
      "5\n",
      "8\n",
      "3\n",
      "2\n",
      "1\n",
      "1\n",
      "8\n",
      "5\n",
      "1\n",
      "2\n",
      "3\n",
      "# the current counter is 300\n",
      "3\n",
      "5\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "3\n",
      "7\n",
      "1\n",
      "3\n",
      "8\n",
      "8\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "0\n",
      "0\n",
      "1\n",
      "1\n",
      "2\n",
      "3\n",
      "3\n",
      "1\n",
      "1\n",
      "2\n",
      "3\n",
      "0\n",
      "1\n",
      "1\n",
      "1\n",
      "3\n",
      "3\n",
      "2\n",
      "3\n",
      "1\n",
      "1\n",
      "1\n",
      "3\n",
      "3\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "2\n",
      "1\n",
      "1\n",
      "3\n",
      "3\n",
      "1\n",
      "2\n",
      "1\n",
      "0\n",
      "1\n",
      "2\n",
      "1\n",
      "1\n",
      "0\n",
      "# the current counter is 360\n",
      "1\n",
      "1\n",
      "0\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "0\n",
      "1\n",
      "2\n",
      "1\n",
      "1\n",
      "2\n",
      "5\n",
      "2\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "8\n",
      "3\n",
      "1\n",
      "3\n",
      "1\n",
      "3\n",
      "1\n",
      "1\n",
      "1\n",
      "0\n",
      "1\n",
      "0\n",
      "1\n",
      "1\n",
      "3\n",
      "3\n",
      "5\n",
      "5\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "2\n",
      "2\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "7\n",
      "3\n",
      "1\n",
      "1\n",
      "# the current counter is 420\n",
      "3\n",
      "1\n",
      "2\n",
      "0\n",
      "1\n",
      "2\n",
      "1\n",
      "3\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "0\n",
      "1\n",
      "1\n",
      "1\n",
      "2\n",
      "7\n",
      "1\n",
      "8\n",
      "5\n",
      "8\n",
      "0\n",
      "1\n",
      "1\n",
      "8\n",
      "0\n",
      "1\n",
      "0\n",
      "1\n",
      "0\n",
      "0\n",
      "1\n",
      "1\n",
      "0\n",
      "1\n",
      "1\n",
      "0\n",
      "1\n",
      "0\n",
      "1\n",
      "3\n",
      "1\n",
      "0\n",
      "2\n",
      "0\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "0\n",
      "1\n",
      "1\n",
      "0\n",
      "1\n",
      "1\n",
      "5\n",
      "2\n",
      "3\n",
      "# the current counter is 480\n",
      "1\n",
      "5\n",
      "3\n",
      "7\n",
      "2\n",
      "0\n",
      "0\n",
      "0\n",
      "1\n",
      "1\n",
      "0\n",
      "1\n",
      "1\n",
      "1\n",
      "2\n",
      "1\n",
      "0\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "3\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "2\n",
      "1\n",
      "1\n",
      "2\n",
      "3\n",
      "1\n",
      "2\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "0\n",
      "1\n",
      "1\n",
      "0\n",
      "0\n",
      "1\n",
      "0\n",
      "3\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "2\n",
      "0\n",
      "2\n",
      "# the current counter is 540\n",
      "2\n",
      "0\n",
      "0\n",
      "1\n",
      "0\n",
      "1\n",
      "3\n",
      "1\n",
      "1\n",
      "8\n",
      "1\n",
      "5\n",
      "9\n",
      "8\n",
      "3\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "0\n",
      "1\n",
      "1\n",
      "2\n",
      "1\n",
      "0\n",
      "1\n",
      "0\n",
      "1\n",
      "2\n",
      "0\n",
      "0\n",
      "1\n",
      "1\n",
      "2\n",
      "1\n",
      "1\n",
      "3\n",
      "7\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "3\n",
      "1\n",
      "3\n",
      "7\n",
      "8\n",
      "8\n",
      "1\n",
      "1\n",
      "8\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "# the current counter is 600\n",
      "2\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "7\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "0\n",
      "1\n",
      "1\n",
      "1\n",
      "2\n",
      "1\n",
      "1\n",
      "1\n",
      "6\n",
      "1\n",
      "1\n",
      "0\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "8\n",
      "3\n",
      "3\n",
      "3\n",
      "1\n",
      "1\n",
      "2\n",
      "1\n",
      "8\n",
      "2\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "0\n",
      "3\n",
      "2\n",
      "2\n",
      "1\n",
      "1\n",
      "3\n",
      "# the current counter is 660\n",
      "8\n",
      "1\n",
      "1\n",
      "2\n",
      "3\n",
      "0\n",
      "1\n",
      "8\n",
      "1\n",
      "1\n",
      "7\n",
      "3\n",
      "1\n",
      "1\n",
      "3\n",
      "3\n",
      "3\n",
      "0\n",
      "1\n",
      "3\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "0\n",
      "1\n",
      "0\n",
      "0\n",
      "1\n",
      "1\n",
      "1\n",
      "0\n",
      "5\n",
      "2\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "2\n",
      "2\n",
      "1\n",
      "2\n",
      "1\n",
      "1\n",
      "0\n",
      "1\n",
      "1\n",
      "1\n",
      "2\n",
      "3\n",
      "1\n",
      "1\n",
      "3\n",
      "1\n",
      "1\n",
      "2\n",
      "1\n",
      "1\n",
      "# the current counter is 720\n",
      "3\n",
      "1\n",
      "1\n",
      "5\n",
      "3\n",
      "1\n",
      "3\n",
      "1\n",
      "1\n",
      "0\n",
      "1\n",
      "1\n",
      "2\n",
      "1\n",
      "2\n",
      "1\n",
      "1\n",
      "2\n",
      "1\n",
      "2\n",
      "1\n",
      "3\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "0\n",
      "0\n",
      "1\n",
      "1\n",
      "0\n",
      "1\n",
      "1\n",
      "0\n",
      "0\n",
      "1\n",
      "1\n",
      "1\n",
      "0\n",
      "1\n",
      "1\n",
      "0\n",
      "0\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "0\n",
      "0\n",
      "0\n",
      "1\n",
      "0\n",
      "1\n",
      "1\n",
      "1\n",
      "3\n",
      "8\n",
      "1\n",
      "# the current counter is 780\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "2\n",
      "5\n",
      "1\n",
      "1\n",
      "2\n",
      "1\n",
      "1\n",
      "2\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "3\n",
      "2\n",
      "3\n",
      "1\n",
      "2\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "2\n",
      "8\n",
      "1\n",
      "1\n",
      "1\n",
      "2\n",
      "1\n",
      "0\n",
      "1\n",
      "1\n",
      "2\n",
      "1\n",
      "3\n",
      "7\n",
      "3\n",
      "7\n",
      "2\n",
      "1\n",
      "1\n",
      "3\n",
      "1\n",
      "1\n",
      "1\n",
      "3\n",
      "3\n",
      "1\n",
      "1\n",
      "3\n",
      "2\n",
      "1\n",
      "1\n",
      "0\n",
      "3\n",
      "# the current counter is 840\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "2\n",
      "2\n",
      "3\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "3\n",
      "3\n",
      "2\n",
      "5\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "2\n",
      "1\n",
      "0\n",
      "2\n",
      "1\n",
      "2\n",
      "3\n",
      "1\n",
      "2\n",
      "1\n",
      "5\n",
      "1\n",
      "3\n",
      "1\n",
      "3\n",
      "3\n",
      "2\n",
      "1\n",
      "3\n",
      "1\n",
      "7\n",
      "1\n",
      "2\n",
      "1\n",
      "1\n",
      "1\n",
      "# the current counter is 900\n",
      "3\n",
      "2\n",
      "2\n",
      "1\n",
      "1\n",
      "1\n",
      "3\n",
      "3\n",
      "5\n",
      "1\n",
      "1\n",
      "0\n",
      "7\n",
      "0\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "3\n",
      "1\n",
      "0\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "0\n",
      "0\n",
      "1\n",
      "2\n",
      "1\n",
      "3\n",
      "0\n",
      "0\n",
      "1\n",
      "2\n",
      "1\n",
      "1\n",
      "3\n",
      "1\n",
      "2\n",
      "1\n",
      "1\n",
      "1\n",
      "8\n",
      "3\n",
      "3\n",
      "3\n",
      "3\n",
      "5\n",
      "1\n",
      "1\n",
      "2\n",
      "1\n",
      "1\n",
      "1\n",
      "0\n",
      "# the current counter is 960\n",
      "1\n",
      "1\n",
      "3\n",
      "1\n",
      "1\n",
      "2\n",
      "1\n",
      "1\n",
      "2\n",
      "1\n",
      "0\n",
      "1\n",
      "0\n",
      "1\n",
      "1\n",
      "2\n",
      "0\n",
      "1\n",
      "0\n",
      "3\n",
      "1\n",
      "0\n",
      "0\n",
      "1\n",
      "1\n",
      "2\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "0\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "0\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "3\n",
      "3\n",
      "1\n",
      "1\n",
      "8\n",
      "0\n",
      "1\n",
      "7\n",
      "1\n",
      "3\n",
      "5\n",
      "0\n",
      "3\n",
      "# the current counter is 1020\n",
      "1\n",
      "0\n",
      "2\n",
      "1\n",
      "1\n",
      "1\n",
      "0\n",
      "1\n",
      "1\n",
      "1\n",
      "0\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "0\n",
      "3\n",
      "1\n",
      "0\n",
      "1\n",
      "1\n",
      "0\n",
      "0\n",
      "0\n",
      "0\n",
      "1\n",
      "0\n",
      "0\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "0\n",
      "1\n",
      "1\n",
      "0\n",
      "2\n",
      "1\n",
      "1\n",
      "1\n",
      "0\n",
      "1\n",
      "1\n",
      "1\n",
      "1\n",
      "0\n",
      "1\n",
      "0\n",
      "1\n",
      "1\n",
      "all 1071 prompts took: 666.963464975357 seconds\n"
     ]
    }
   ],
   "source": [
    "# if the column mistral_medium_ratings already exists, then read skip this execution\n",
    "if \"mistral_medium_ratings\" in list(experiment_df.columns):\n",
    "    print(\"mistral_medium_ratings already exists\")\n",
    "else:\n",
    "    model = \"mistral-medium\"\n",
    "    # get the score for each prompt\n",
    "    mistral_medium_responses = []\n",
    "    mistral_medium_response_objects = []\n",
    "    start_timestamp = time.time()\n",
    "    last_start_timestamp = time.time()\n",
    "    counter = 0\n",
    "    for prompt in verses:\n",
    "        counter += 1\n",
    "        if counter%60 == 0:\n",
    "            print(\"# the current counter is\", counter)\n",
    "        response = get_mistral_response(model, prompt)\n",
    "        print(response.choices[0].message.content)\n",
    "        mistral_medium_responses.append(response.choices[0].message.content)\n",
    "        mistral_medium_response_objects.append(response)\n",
    "    \n",
    "    total_time_for_mistral_medium = time.time() - start_timestamp\n",
    "    print(\"all\", len(verses), \"prompts took:\", total_time_for_mistral_medium, \"seconds\")\n",
    "    \n",
    "    mistral_medium_responses_int = [convert_to_int(response.strip()) if response.strip() != \"\" else -3 for response in mistral_medium_responses ]\n",
    "\n",
    "    experiment_df.insert(2,\"mistral_medium_ratings\", mistral_medium_responses_int)\n",
    "    experiment_df.insert(2,\"mistral_medium_response_objects\", mistral_medium_response_objects)\n",
    "    \n",
    "    # save time in times_df\n",
    "    times_df = pd.concat([times_df, pd.DataFrame([{'name': 'mistral_medium', 'sec': total_time_for_mistral_medium}])])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 60,
   "metadata": {},
   "outputs": [],
   "source": [
    "experiment_df.to_csv(\"data/experiment_df.csv\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 59,
   "metadata": {},
   "outputs": [],
   "source": [
    "experiment_df.to_json(\"data/experiment_df.json\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 63,
   "metadata": {},
   "outputs": [],
   "source": [
    "times_df.to_csv(\"data/times_df.csv\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 74,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>name</th>\n",
       "      <th>sec</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>text_bison</td>\n",
       "      <td>549.711793</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>gemini_pro</td>\n",
       "      <td>456.114870</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>gpt_3_5</td>\n",
       "      <td>2752.802603</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>gpt_4</td>\n",
       "      <td>1070.960249</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>ollama_mistral7B</td>\n",
       "      <td>1201.558121</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>ollama_llama7B</td>\n",
       "      <td>1091.680313</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>ollama_llama13B</td>\n",
       "      <td>3446.577645</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>mistral_tiny</td>\n",
       "      <td>3814.597317</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>mistral_small</td>\n",
       "      <td>908.788969</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>mistral_medium</td>\n",
       "      <td>666.963465</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "               name          sec\n",
       "0  text_bison        549.711793 \n",
       "1  gemini_pro        456.114870 \n",
       "2  gpt_3_5           2752.802603\n",
       "3  gpt_4             1070.960249\n",
       "4  ollama_mistral7B  1201.558121\n",
       "5  ollama_llama7B    1091.680313\n",
       "6  ollama_llama13B   3446.577645\n",
       "7  mistral_tiny      3814.597317\n",
       "8  mistral_small     908.788969 \n",
       "9  mistral_medium    666.963465 "
      ]
     },
     "execution_count": 74,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "times_df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 88,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1000x600 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# compare how long each model took to process all the prompts\n",
    "# make a bar plot\n",
    "plt.figure(figsize=(10, 6))  # Adjust the figure size\n",
    "plt.bar(times_df.name, times_df.sec, color='orange', width=0.5)  # Adjust bar width and add colors\n",
    "plt.ylabel(\"Time in seconds\")\n",
    "plt.title(\"Time to produce all ratings for all models in seconds\")\n",
    "\n",
    "# Add data labels above the bars\n",
    "for i, el in enumerate(np.round(times_df.sec)):\n",
    "    plt.text(i, el + 10, f\"{el:.0f}\", ha='center', fontsize=12)\n",
    "\n",
    "plt.ylim(0, max(times_df.sec) + 200)  # Adjust the y-axis limit for better visualization\n",
    "\n",
    "# Rotate x-axis labels for better readability\n",
    "plt.xticks(ha='left', fontsize=10, rotation=-20, wrap=True)\n",
    "\n",
    "# Add a grid for better reference\n",
    "plt.grid(axis='y', linestyle='--', alpha=0.7)\n",
    "\n",
    "# Save the plot as \"costs.png\" with high DPI for better quality\n",
    "plt.savefig(\"images/times.png\", dpi=300)\n",
    "plt.tight_layout()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Analyze reliability\n",
    "The rating is set to less than 0 if something failed.  \n",
    "Lets count all the cells where the rating is less than 0 and visualize the numbers in a barplot."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['gpt_4_ratings',\n",
       " 'mistral_medium_ratings',\n",
       " 'mistral_small_ratings',\n",
       " 'mistral_tiny_ratings',\n",
       " 'ollama_llama13B_ratings',\n",
       " 'ollama_llama7B_ratings',\n",
       " 'ollama_mistral7B_ratings',\n",
       " 'gpt_3_5_ratings',\n",
       " 'gemini_pro_ratings',\n",
       " 'text_bison_ratings']"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "ratings_columns = [col for col in experiment_df.columns if \"ratings\" in col]\n",
    "ratings_columns"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'gpt_4_ratings': 0,\n",
       " 'mistral_medium_ratings': 0,\n",
       " 'mistral_small_ratings': 20,\n",
       " 'mistral_tiny_ratings': 2,\n",
       " 'ollama_llama13B_ratings': 495,\n",
       " 'ollama_llama7B_ratings': 0,\n",
       " 'ollama_mistral7B_ratings': 3,\n",
       " 'gpt_3_5_ratings': 0,\n",
       " 'gemini_pro_ratings': 31,\n",
       " 'text_bison_ratings': 0}"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# count the cells with ratings less than 0\n",
    "counts = {}\n",
    "for col in ratings_columns:\n",
    "    counts[col] = experiment_df[col].map(lambda el: 1 if el < 0 else 0).sum()\n",
    "counts"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1000x600 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# make a barplot for this data\n",
    "plt.figure(figsize=(10, 6))  # Adjust the figure size\n",
    "labels = [col.removesuffix(\"_ratings\").replace(\"_\", \" \") for col in ratings_columns]\n",
    "plt.bar(labels, counts.values(), color='orange', width=0.5)  # Adjust bar width and add colors\n",
    "\n",
    "# add a data label above the bars\n",
    "for i, el in enumerate(counts.values()):\n",
    "    plt.text(i, el + 5, f\"{el}\", ha='center', fontsize=12)\n",
    "\n",
    "plt.ylim(0, max(counts.values()) + 50)  # Adjust the y-axis limit for better visualization\n",
    "\n",
    "plt.xticks(ha='left', fontsize=10, rotation=-20, wrap=True)\n",
    "plt.ylabel(\"Number of errors\")\n",
    "plt.title(\"Number of errors for each model\")\n",
    "\n",
    "# save the plot\n",
    "plt.savefig(\"images/errors.png\", dpi=300)\n",
    "plt.tight_layout()\n",
    "plt.show()"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.10.12"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}