mixture-of-experts.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [],
   "source": [
    "from datetime import datetime\n",
    "import json\n",
    "from utils import load_all_llm_answers_from_json\n",
    "import pandas as pd"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## TO DO\n",
    "- HEATMAP of each model and whether they get the same questions wrong or not\n",
    "- Add probabilistic outputs to each model answer\n",
    "- Add \"abstain\" option for all models"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'claude-3-5-sonnet-20240620':      level_0  index         category  \\\n",
       " 0          0      1           Puzzle   \n",
       " 1          1      2           Puzzle   \n",
       " 2          2      3          Spatial   \n",
       " 3          3      4           Puzzle   \n",
       " 4          4      5           Puzzle   \n",
       " ..       ...    ...              ...   \n",
       " 295      295     26           Puzzle   \n",
       " 296      296     27           Puzzle   \n",
       " 297      297     28  Popular science   \n",
       " 298      298     29       Relational   \n",
       " 299      299     30  Popular science   \n",
       " \n",
       "                                               question  \\\n",
       " 0    You have six horses and want to race them to s...   \n",
       " 1    Suppose you're on a game show, and you're give...   \n",
       " 2    You are playing Russian roulette with a six-sh...   \n",
       " 3    A farmer wants to cross a river and take with ...   \n",
       " 4    Bob has three boxes in front of him - Box A, B...   \n",
       " ..                                                 ...   \n",
       " 295  A runaway trolley is heading down the tracks a...   \n",
       " 296  How do you efficiently measure exactly 4 gallo...   \n",
       " 297  A 2kg tree grows in a planted pot with 10kg of...   \n",
       " 298  A man steps over a 0.5m hurdle. What is the mi...   \n",
       " 299  I walk to my friend's house, averaging 3mph. H...   \n",
       " \n",
       "                                           human_answer  \\\n",
       " 0    Race them on a single race track with at least...   \n",
       " 1    It is not an advantage to switch. It makes no ...   \n",
       " 2    Yes, you should spin again. There was only one...   \n",
       " 3    Place the wolf, goat, and cabbage in separate ...   \n",
       " 4    Bob should not open a box because he has a 1/3...   \n",
       " ..                                                 ...   \n",
       " 295  It doesn't impact anyones lives, as the trolle...   \n",
       " 296                           Fill up the 4-gallon jug   \n",
       " 297  There is still 10kg of soil. Trees are primari...   \n",
       " 298  0m, he didn't jump at all, he stepped over the...   \n",
       " 299  Instantaneously fast, it is mathematically imp...   \n",
       " \n",
       "                              correct_answer  \\\n",
       " 0                                       One   \n",
       " 1    They have equal probability of winning   \n",
       " 2                Yes, you should spin again   \n",
       " 3                              One crossing   \n",
       " 4           No, Bob should not open any box   \n",
       " ..                                      ...   \n",
       " 295          It doesn't impact their lives.   \n",
       " 296              Another option not stated.   \n",
       " 297                                    10kg   \n",
       " 298                                      0m   \n",
       " 299                         Infinitely fast   \n",
       " \n",
       "                                        multiple_choice  \\\n",
       " 0                           [Three, One, Two, Unknown]   \n",
       " 1    [Door No.3, Door No.1, Door No.2, They have eq...   \n",
       " 2    [The probabilities are random and unknown, It ...   \n",
       " 3    [The animals are not able to be transported ac...   \n",
       " 4    [Yes, Box A, Yes, Box C, Yes, Box B, No, Bob s...   \n",
       " ..                                                 ...   \n",
       " 295  [Pulling the lever will take their lives., Pul...   \n",
       " 296  [Another option not stated., Filling the 2-gal...   \n",
       " 297                             [9kg, 10kg, 11kg, 8kg]   \n",
       " 298  [0.5m, 0m, Slightly higher than 0.5m, Between ...   \n",
       " 299               [6mph, 9mph, Infinitely fast, 12mph]   \n",
       " \n",
       "                                  multi_choice_question correct_letter  \\\n",
       " 0    QUESTION\\nYou have six horses and want to race...              B   \n",
       " 1    QUESTION\\nSuppose you're on a game show, and y...              D   \n",
       " 2    QUESTION\\nYou are playing Russian roulette wit...              C   \n",
       " 3    QUESTION\\nA farmer wants to cross a river and ...              D   \n",
       " 4    QUESTION\\nBob has three boxes in front of him ...              D   \n",
       " ..                                                 ...            ...   \n",
       " 295  QUESTION\\nA runaway trolley is heading down th...              C   \n",
       " 296  QUESTION\\nHow do you efficiently measure exact...              A   \n",
       " 297  QUESTION\\nA 2kg tree grows in a planted pot wi...              B   \n",
       " 298  QUESTION\\nA man steps over a 0.5m hurdle. What...              B   \n",
       " 299  QUESTION\\nI walk to my friend's house, averagi...              C   \n",
       " \n",
       "                                           model_answer  score  \\\n",
       " 0    Let's think through this step-by-step:\\n\\n1. W...      0   \n",
       " 1    Let's think through this step-by-step:\\n\\n1. I...    100   \n",
       " 2    Let's think through this step-by-step:\\n\\n1. I...    100   \n",
       " 3    Let's think through this step-by-step:\\n\\n1. T...    100   \n",
       " 4    Let's think through this logically:\\n\\n1. Bob ...    100   \n",
       " ..                                                 ...    ...   \n",
       " 295  I'll analyze this scenario and provide my reas...    100   \n",
       " 296  Let's think through this step-by-step:\\n\\n1. W...      0   \n",
       " 297  Let's think through this step-by-step:\\n\\n1. I...      0   \n",
       " 298  Let's think through this step-by-step:\\n\\n1. T...      0   \n",
       " 299  Let's think through this step-by-step:\\n\\n1. L...    100   \n",
       " \n",
       "          json_answer json_answer_letter  invalid_answer_letter  \n",
       " 0    {'ANSWER': 'C'}                  C                      0  \n",
       " 1    {'ANSWER': 'D'}                  D                      0  \n",
       " 2    {'ANSWER': 'C'}                  C                      0  \n",
       " 3    {'ANSWER': 'D'}                  D                      0  \n",
       " 4    {'ANSWER': 'D'}                  D                      0  \n",
       " ..               ...                ...                    ...  \n",
       " 295  {'ANSWER': 'C'}                  C                      0  \n",
       " 296  {'ANSWER': 'B'}                  B                      0  \n",
       " 297  {'ANSWER': 'A'}                  A                      0  \n",
       " 298  {'ANSWER': 'C'}                  C                      0  \n",
       " 299  {'ANSWER': 'C'}                  C                      0  \n",
       " \n",
       " [300 rows x 14 columns],\n",
       " 'claude-3-opus-20240229':      level_0  index         category  \\\n",
       " 0          0      1           Puzzle   \n",
       " 1          1      2           Puzzle   \n",
       " 2          2      3          Spatial   \n",
       " 3          3      4           Puzzle   \n",
       " 4          4      5           Puzzle   \n",
       " ..       ...    ...              ...   \n",
       " 295      295     26           Puzzle   \n",
       " 296      296     27           Puzzle   \n",
       " 297      297     28  Popular science   \n",
       " 298      298     29       Relational   \n",
       " 299      299     30  Popular science   \n",
       " \n",
       "                                               question  \\\n",
       " 0    You have six horses and want to race them to s...   \n",
       " 1    Suppose you're on a game show, and you're give...   \n",
       " 2    You are playing Russian roulette with a six-sh...   \n",
       " 3    A farmer wants to cross a river and take with ...   \n",
       " 4    Bob has three boxes in front of him - Box A, B...   \n",
       " ..                                                 ...   \n",
       " 295  A runaway trolley is heading down the tracks a...   \n",
       " 296  How do you efficiently measure exactly 4 gallo...   \n",
       " 297  A 2kg tree grows in a planted pot with 10kg of...   \n",
       " 298  A man steps over a 0.5m hurdle. What is the mi...   \n",
       " 299  I walk to my friend's house, averaging 3mph. H...   \n",
       " \n",
       "                                           human_answer  \\\n",
       " 0    Race them on a single race track with at least...   \n",
       " 1    It is not an advantage to switch. It makes no ...   \n",
       " 2    Yes, you should spin again. There was only one...   \n",
       " 3    Place the wolf, goat, and cabbage in separate ...   \n",
       " 4    Bob should not open a box because he has a 1/3...   \n",
       " ..                                                 ...   \n",
       " 295  It doesn't impact anyones lives, as the trolle...   \n",
       " 296                           Fill up the 4-gallon jug   \n",
       " 297  There is still 10kg of soil. Trees are primari...   \n",
       " 298  0m, he didn't jump at all, he stepped over the...   \n",
       " 299  Instantaneously fast, it is mathematically imp...   \n",
       " \n",
       "                              correct_answer  \\\n",
       " 0                                       One   \n",
       " 1    They have equal probability of winning   \n",
       " 2                Yes, you should spin again   \n",
       " 3                              One crossing   \n",
       " 4           No, Bob should not open any box   \n",
       " ..                                      ...   \n",
       " 295          It doesn't impact their lives.   \n",
       " 296              Another option not stated.   \n",
       " 297                                    10kg   \n",
       " 298                                      0m   \n",
       " 299                         Infinitely fast   \n",
       " \n",
       "                                        multiple_choice  \\\n",
       " 0                           [Three, One, Two, Unknown]   \n",
       " 1    [Door No.3, Door No.1, Door No.2, They have eq...   \n",
       " 2    [The probabilities are random and unknown, It ...   \n",
       " 3    [The animals are not able to be transported ac...   \n",
       " 4    [Yes, Box A, Yes, Box C, Yes, Box B, No, Bob s...   \n",
       " ..                                                 ...   \n",
       " 295  [Pulling the lever will take their lives., Pul...   \n",
       " 296  [Another option not stated., Filling the 2-gal...   \n",
       " 297                             [9kg, 10kg, 11kg, 8kg]   \n",
       " 298  [0.5m, 0m, Slightly higher than 0.5m, Between ...   \n",
       " 299               [6mph, 9mph, Infinitely fast, 12mph]   \n",
       " \n",
       "                                  multi_choice_question correct_letter  \\\n",
       " 0    QUESTION\\nYou have six horses and want to race...              B   \n",
       " 1    QUESTION\\nSuppose you're on a game show, and y...              D   \n",
       " 2    QUESTION\\nYou are playing Russian roulette wit...              C   \n",
       " 3    QUESTION\\nA farmer wants to cross a river and ...              D   \n",
       " 4    QUESTION\\nBob has three boxes in front of him ...              D   \n",
       " ..                                                 ...            ...   \n",
       " 295  QUESTION\\nA runaway trolley is heading down th...              C   \n",
       " 296  QUESTION\\nHow do you efficiently measure exact...              A   \n",
       " 297  QUESTION\\nA 2kg tree grows in a planted pot wi...              B   \n",
       " 298  QUESTION\\nA man steps over a 0.5m hurdle. What...              B   \n",
       " 299  QUESTION\\nI walk to my friend's house, averagi...              C   \n",
       " \n",
       "                                           model_answer  score  \\\n",
       " 0    Here is my explanation and answer in JSON form...      0   \n",
       " 1    Here is my explanation and answer in JSON form...      0   \n",
       " 2    Here is my explanation and answer in JSON form...    100   \n",
       " 3    Here is my explanation and answer in JSON form...      0   \n",
       " 4    Here is my analysis:\\n\\nBob should open a box,...      0   \n",
       " ..                                                 ...    ...   \n",
       " 295  Here is my analysis of the trolley problem:\\n\\...      0   \n",
       " 296  Here is my explanation and answer in JSON form...    100   \n",
       " 297  Here's my explanation and answer in JSON forma...    100   \n",
       " 298  Here is my explanation and answer in JSON form...      0   \n",
       " 299  To solve this problem, let's break it down ste...    100   \n",
       " \n",
       "                                            json_answer json_answer_letter  \\\n",
       " 0                                      {'ANSWER': 'A'}                  A   \n",
       " 1                                      {'ANSWER': 'A'}                  A   \n",
       " 2    {'EXPLANATION': \"If your opponent put in 5 bul...                  C   \n",
       " 3                                      {'ANSWER': 'C'}                  C   \n",
       " 4                                      {'ANSWER': 'C'}                  C   \n",
       " ..                                                 ...                ...   \n",
       " 295                                    {'ANSWER': 'B'}                  B   \n",
       " 296  {'EXPLANATION': 'The most efficient way to mea...                  A   \n",
       " 297  {'EXPLANATION': \"The question states that the ...                  B   \n",
       " 298                                    {'ANSWER': 'C'}                  C   \n",
       " 299                                    {'ANSWER': 'C'}                  C   \n",
       " \n",
       "      invalid_answer_letter  \n",
       " 0                        0  \n",
       " 1                        0  \n",
       " 2                        0  \n",
       " 3                        0  \n",
       " 4                        0  \n",
       " ..                     ...  \n",
       " 295                      0  \n",
       " 296                      0  \n",
       " 297                      0  \n",
       " 298                      0  \n",
       " 299                      0  \n",
       " \n",
       " [300 rows x 14 columns],\n",
       " 'command-r-plus':      level_0  index         category  \\\n",
       " 0          0      1           Puzzle   \n",
       " 1          1      2           Puzzle   \n",
       " 2          2      3          Spatial   \n",
       " 3          3      4           Puzzle   \n",
       " 4          4      5           Puzzle   \n",
       " ..       ...    ...              ...   \n",
       " 295      295     26           Puzzle   \n",
       " 296      296     27           Puzzle   \n",
       " 297      297     28  Popular science   \n",
       " 298      298     29       Relational   \n",
       " 299      299     30  Popular science   \n",
       " \n",
       "                                               question  \\\n",
       " 0    You have six horses and want to race them to s...   \n",
       " 1    Suppose you're on a game show, and you're give...   \n",
       " 2    You are playing Russian roulette with a six-sh...   \n",
       " 3    A farmer wants to cross a river and take with ...   \n",
       " 4    Bob has three boxes in front of him - Box A, B...   \n",
       " ..                                                 ...   \n",
       " 295  A runaway trolley is heading down the tracks a...   \n",
       " 296  How do you efficiently measure exactly 4 gallo...   \n",
       " 297  A 2kg tree grows in a planted pot with 10kg of...   \n",
       " 298  A man steps over a 0.5m hurdle. What is the mi...   \n",
       " 299  I walk to my friend's house, averaging 3mph. H...   \n",
       " \n",
       "                                           human_answer  \\\n",
       " 0    Race them on a single race track with at least...   \n",
       " 1    It is not an advantage to switch. It makes no ...   \n",
       " 2    Yes, you should spin again. There was only one...   \n",
       " 3    Place the wolf, goat, and cabbage in separate ...   \n",
       " 4    Bob should not open a box because he has a 1/3...   \n",
       " ..                                                 ...   \n",
       " 295  It doesn't impact anyones lives, as the trolle...   \n",
       " 296                           Fill up the 4-gallon jug   \n",
       " 297  There is still 10kg of soil. Trees are primari...   \n",
       " 298  0m, he didn't jump at all, he stepped over the...   \n",
       " 299  Instantaneously fast, it is mathematically imp...   \n",
       " \n",
       "                              correct_answer  \\\n",
       " 0                                       One   \n",
       " 1    They have equal probability of winning   \n",
       " 2                Yes, you should spin again   \n",
       " 3                              One crossing   \n",
       " 4           No, Bob should not open any box   \n",
       " ..                                      ...   \n",
       " 295          It doesn't impact their lives.   \n",
       " 296              Another option not stated.   \n",
       " 297                                    10kg   \n",
       " 298                                      0m   \n",
       " 299                         Infinitely fast   \n",
       " \n",
       "                                        multiple_choice  \\\n",
       " 0                           [Two, One, Three, Unknown]   \n",
       " 1    [Door No.1, They have equal probability of win...   \n",
       " 2    [Yes, you should spin again, No, you should no...   \n",
       " 3    [One crossing, The animals are not able to be ...   \n",
       " 4    [Yes, Box A, No, Bob should not open any box, ...   \n",
       " ..                                                 ...   \n",
       " 295  [This is an ethical dilemma., It doesn't impac...   \n",
       " 296  [Another option not stated., Filling the 5-gal...   \n",
       " 297                             [10kg, 11kg, 8kg, 9kg]   \n",
       " 298  [0m, Slightly higher than 0.5m, Between 0.1m a...   \n",
       " 299               [9mph, Infinitely fast, 6mph, 12mph]   \n",
       " \n",
       "                                  multi_choice_question correct_letter  \\\n",
       " 0    QUESTION\\nYou have six horses and want to race...              B   \n",
       " 1    QUESTION\\nSuppose you're on a game show, and y...              B   \n",
       " 2    QUESTION\\nYou are playing Russian roulette wit...              A   \n",
       " 3    QUESTION\\nA farmer wants to cross a river and ...              A   \n",
       " 4    QUESTION\\nBob has three boxes in front of him ...              B   \n",
       " ..                                                 ...            ...   \n",
       " 295  QUESTION\\nA runaway trolley is heading down th...              B   \n",
       " 296  QUESTION\\nHow do you efficiently measure exact...              A   \n",
       " 297  QUESTION\\nA 2kg tree grows in a planted pot wi...              A   \n",
       " 298  QUESTION\\nA man steps over a 0.5m hurdle. What...              A   \n",
       " 299  QUESTION\\nI walk to my friend's house, averagi...              B   \n",
       " \n",
       "                                           model_answer  score  \\\n",
       " 0    Solution:\\nIf there are 6 horses, we can have ...    100   \n",
       " 1    {\"ANSWER\": \"D\"}\\n\\nThis is a probability puzzl...      0   \n",
       " 2    {\"ANSWER\": \"B\"}\\n\\nExplanation: If your oppone...      0   \n",
       " 3    Explanation: This is a variation of the classi...      0   \n",
       " 4    {\"ANSWER\": \"B\"}\\n\\nExplanation: Bob should not...    100   \n",
       " ..                                                 ...    ...   \n",
       " 295  {\\n    \"ANSWER\": \"C\"\\n}\\n\\nExplanation: This i...      0   \n",
       " 296  ```json\\n{\\n    \"ANSWER\": \"A\"\\n}\\n```\\n\\n**Exp...    100   \n",
       " 297  ```json\\n{\\n    \"ANSWER\": \"D\"\\n}\\n```\\n\\n**Exp...      0   \n",
       " 298  ```json\\n{\\n    \"ANSWER\": \"B\"\\n}\\n```\\n\\n**Exp...      0   \n",
       " 299  Solution:\\nLet X be the running speed.\\n\\nThe ...      0   \n",
       " \n",
       "          json_answer json_answer_letter  invalid_answer_letter  \n",
       " 0    {'ANSWER': 'B'}                  B                      0  \n",
       " 1    {'ANSWER': 'D'}                  D                      0  \n",
       " 2    {'ANSWER': 'B'}                  B                      0  \n",
       " 3    {'ANSWER': 'C'}                  C                      0  \n",
       " 4    {'ANSWER': 'B'}                  B                      0  \n",
       " ..               ...                ...                    ...  \n",
       " 295  {'ANSWER': 'C'}                  C                      0  \n",
       " 296  {'ANSWER': 'A'}                  A                      0  \n",
       " 297  {'ANSWER': 'D'}                  D                      0  \n",
       " 298  {'ANSWER': 'B'}                  B                      0  \n",
       " 299  {'ANSWER': 'A'}                  A                      0  \n",
       " \n",
       " [300 rows x 14 columns],\n",
       " 'gemini-1_5-pro':      level_0  index         category  \\\n",
       " 0          0      1           Puzzle   \n",
       " 1          1      2           Puzzle   \n",
       " 2          2      3          Spatial   \n",
       " 3          3      4           Puzzle   \n",
       " 4          4      5           Puzzle   \n",
       " ..       ...    ...              ...   \n",
       " 295      295     26           Puzzle   \n",
       " 296      296     27           Puzzle   \n",
       " 297      297     28  Popular science   \n",
       " 298      298     29       Relational   \n",
       " 299      299     30  Popular science   \n",
       " \n",
       "                                               question  \\\n",
       " 0    You have six horses and want to race them to s...   \n",
       " 1    Suppose you're on a game show, and you're give...   \n",
       " 2    You are playing Russian roulette with a six-sh...   \n",
       " 3    A farmer wants to cross a river and take with ...   \n",
       " 4    Bob has three boxes in front of him - Box A, B...   \n",
       " ..                                                 ...   \n",
       " 295  A runaway trolley is heading down the tracks a...   \n",
       " 296  How do you efficiently measure exactly 4 gallo...   \n",
       " 297  A 2kg tree grows in a planted pot with 10kg of...   \n",
       " 298  A man steps over a 0.5m hurdle. What is the mi...   \n",
       " 299  I walk to my friend's house, averaging 3mph. H...   \n",
       " \n",
       "                                           human_answer  \\\n",
       " 0    Race them on a single race track with at least...   \n",
       " 1    It is not an advantage to switch. It makes no ...   \n",
       " 2    Yes, you should spin again. There was only one...   \n",
       " 3    Place the wolf, goat, and cabbage in separate ...   \n",
       " 4    Bob should not open a box because he has a 1/3...   \n",
       " ..                                                 ...   \n",
       " 295  It doesn't impact anyones lives, as the trolle...   \n",
       " 296                           Fill up the 4-gallon jug   \n",
       " 297  There is still 10kg of soil. Trees are primari...   \n",
       " 298  0m, he didn't jump at all, he stepped over the...   \n",
       " 299  Instantaneously fast, it is mathematically imp...   \n",
       " \n",
       "                              correct_answer  \\\n",
       " 0                                       One   \n",
       " 1    They have equal probability of winning   \n",
       " 2                Yes, you should spin again   \n",
       " 3                              One crossing   \n",
       " 4           No, Bob should not open any box   \n",
       " ..                                      ...   \n",
       " 295          It doesn't impact their lives.   \n",
       " 296              Another option not stated.   \n",
       " 297                                    10kg   \n",
       " 298                                      0m   \n",
       " 299                         Infinitely fast   \n",
       " \n",
       "                                        multiple_choice  \\\n",
       " 0                           [Three, One, Two, Unknown]   \n",
       " 1    [Door No.3, Door No.1, Door No.2, They have eq...   \n",
       " 2    [The probabilities are random and unknown, It ...   \n",
       " 3    [The animals are not able to be transported ac...   \n",
       " 4    [Yes, Box A, Yes, Box C, Yes, Box B, No, Bob s...   \n",
       " ..                                                 ...   \n",
       " 295  [Pulling the lever will take their lives., Pul...   \n",
       " 296  [Another option not stated., Filling the 2-gal...   \n",
       " 297                             [9kg, 10kg, 11kg, 8kg]   \n",
       " 298  [0.5m, 0m, Slightly higher than 0.5m, Between ...   \n",
       " 299               [6mph, 9mph, Infinitely fast, 12mph]   \n",
       " \n",
       "                                  multi_choice_question correct_letter  \\\n",
       " 0    QUESTION\\nYou have six horses and want to race...              B   \n",
       " 1    QUESTION\\nSuppose you're on a game show, and y...              D   \n",
       " 2    QUESTION\\nYou are playing Russian roulette wit...              C   \n",
       " 3    QUESTION\\nA farmer wants to cross a river and ...              D   \n",
       " 4    QUESTION\\nBob has three boxes in front of him ...              D   \n",
       " ..                                                 ...            ...   \n",
       " 295  QUESTION\\nA runaway trolley is heading down th...              C   \n",
       " 296  QUESTION\\nHow do you efficiently measure exact...              A   \n",
       " 297  QUESTION\\nA 2kg tree grows in a planted pot wi...              B   \n",
       " 298  QUESTION\\nA man steps over a 0.5m hurdle. What...              B   \n",
       " 299  QUESTION\\nI walk to my friend's house, averagi...              C   \n",
       " \n",
       "                                           model_answer  score  \\\n",
       " 0    ```json\\n{\\n \"ANSWER\": \"C\",\\n \"EXPLANATION\": \"...      0   \n",
       " 1    Here's the breakdown of why the answer is **C....      0   \n",
       " 2    ```json\\n{\\n \"ANSWER\": \"D\",\\n \"Explanation\": \"...      0   \n",
       " 3    ```json\\n{\\n \"ANSWER\": \"C\",\\n \"EXPLANATION\": \"...      0   \n",
       " 4    ```json\\n{\\n \"ANSWER\": \"D\",\\n \"Explanation\": \"...    100   \n",
       " ..                                                 ...    ...   \n",
       " 295  ```json\\n{\\n \"ANSWER\": \"B\",\\n \"Explanation\": \"...      0   \n",
       " 296  ```json\\n{\\n \"ANSWER\": \"B\",\\n \"EXPLANATION\": \"...      0   \n",
       " 297  {\"ANSWER\": \"B\",\\n\"EXPLANATION\": \"The growth of...    100   \n",
       " 298  Here's the breakdown:\\n\\n* **The scenario:** T...      0   \n",
       " 299  ```json\\n{\\n \"ANSWER\": \"C\",\\n \"Explanation\": \"...    100   \n",
       " \n",
       "                                            json_answer json_answer_letter  \\\n",
       " 0    {'ANSWER': 'C', 'EXPLANATION': \"Here's how to ...                  C   \n",
       " 1                                      {'ANSWER': 'C'}                  C   \n",
       " 2    {'ANSWER': 'D', 'Explanation': \"Here's why you...                  D   \n",
       " 3    {'ANSWER': 'C', 'EXPLANATION': \"Here's how the...                  C   \n",
       " 4    {'ANSWER': 'D', 'Explanation': \"Bob has no way...                  D   \n",
       " ..                                                 ...                ...   \n",
       " 295  {'ANSWER': 'B', 'Explanation': \"Pulling the le...                  B   \n",
       " 296  {'ANSWER': 'B', 'EXPLANATION': 'The simplest s...                  B   \n",
       " 297  {'ANSWER': 'B', 'EXPLANATION': \"The growth of ...                  B   \n",
       " 298                                    {'ANSWER': 'C'}                  C   \n",
       " 299  {'ANSWER': 'C', 'Explanation': \"Here's why:  L...                  C   \n",
       " \n",
       "      invalid_answer_letter  \n",
       " 0                        0  \n",
       " 1                        0  \n",
       " 2                        0  \n",
       " 3                        0  \n",
       " 4                        0  \n",
       " ..                     ...  \n",
       " 295                      0  \n",
       " 296                      0  \n",
       " 297                      0  \n",
       " 298                      0  \n",
       " 299                      0  \n",
       " \n",
       " [300 rows x 14 columns],\n",
       " 'gpt-4-turbo-preview':      level_0  index         category  \\\n",
       " 0          0      1           Puzzle   \n",
       " 1          1      2           Puzzle   \n",
       " 2          2      3          Spatial   \n",
       " 3          3      4           Puzzle   \n",
       " 4          4      5           Puzzle   \n",
       " ..       ...    ...              ...   \n",
       " 295      295     26           Puzzle   \n",
       " 296      296     27           Puzzle   \n",
       " 297      297     28  Popular science   \n",
       " 298      298     29       Relational   \n",
       " 299      299     30  Popular science   \n",
       " \n",
       "                                               question  \\\n",
       " 0    You have six horses and want to race them to s...   \n",
       " 1    Suppose you're on a game show, and you're give...   \n",
       " 2    You are playing Russian roulette with a six-sh...   \n",
       " 3    A farmer wants to cross a river and take with ...   \n",
       " 4    Bob has three boxes in front of him - Box A, B...   \n",
       " ..                                                 ...   \n",
       " 295  A runaway trolley is heading down the tracks a...   \n",
       " 296  How do you efficiently measure exactly 4 gallo...   \n",
       " 297  A 2kg tree grows in a planted pot with 10kg of...   \n",
       " 298  A man steps over a 0.5m hurdle. What is the mi...   \n",
       " 299  I walk to my friend's house, averaging 3mph. H...   \n",
       " \n",
       "                                           human_answer  \\\n",
       " 0    Race them on a single race track with at least...   \n",
       " 1    It is not an advantage to switch. It makes no ...   \n",
       " 2    Yes, you should spin again. There was only one...   \n",
       " 3    Place the wolf, goat, and cabbage in separate ...   \n",
       " 4    Bob should not open a box because he has a 1/3...   \n",
       " ..                                                 ...   \n",
       " 295  It doesn't impact anyones lives, as the trolle...   \n",
       " 296                           Fill up the 4-gallon jug   \n",
       " 297  There is still 10kg of soil. Trees are primari...   \n",
       " 298  0m, he didn't jump at all, he stepped over the...   \n",
       " 299  Instantaneously fast, it is mathematically imp...   \n",
       " \n",
       "                              correct_answer  \\\n",
       " 0                                       One   \n",
       " 1    They have equal probability of winning   \n",
       " 2                Yes, you should spin again   \n",
       " 3                              One crossing   \n",
       " 4           No, Bob should not open any box   \n",
       " ..                                      ...   \n",
       " 295          It doesn't impact their lives.   \n",
       " 296              Another option not stated.   \n",
       " 297                                    10kg   \n",
       " 298                                      0m   \n",
       " 299                         Infinitely fast   \n",
       " \n",
       "                                        multiple_choice  \\\n",
       " 0                           [Two, Three, Unknown, One]   \n",
       " 1    [Door No.2, Door No.3, Door No.1, They have eq...   \n",
       " 2    [Yes, you should spin again, No, you should no...   \n",
       " 3    [Five crossings, The animals are not able to b...   \n",
       " 4    [Yes, Box A, No, Bob should not open any box, ...   \n",
       " ..                                                 ...   \n",
       " 295  [Pulling the lever will take their lives., Pul...   \n",
       " 296  [Another option not stated., Filling the 2-gal...   \n",
       " 297                             [9kg, 10kg, 11kg, 8kg]   \n",
       " 298  [0.5m, 0m, Slightly higher than 0.5m, Between ...   \n",
       " 299               [6mph, 9mph, Infinitely fast, 12mph]   \n",
       " \n",
       "                                  multi_choice_question correct_letter  \\\n",
       " 0    QUESTION\\nYou have six horses and want to race...              D   \n",
       " 1    QUESTION\\nSuppose you're on a game show, and y...              D   \n",
       " 2    QUESTION\\nYou are playing Russian roulette wit...              A   \n",
       " 3    QUESTION\\nA farmer wants to cross a river and ...              C   \n",
       " 4    QUESTION\\nBob has three boxes in front of him ...              B   \n",
       " ..                                                 ...            ...   \n",
       " 295  QUESTION\\nA runaway trolley is heading down th...              C   \n",
       " 296  QUESTION\\nHow do you efficiently measure exact...              A   \n",
       " 297  QUESTION\\nA 2kg tree grows in a planted pot wi...              B   \n",
       " 298  QUESTION\\nA man steps over a 0.5m hurdle. What...              B   \n",
       " 299  QUESTION\\nI walk to my friend's house, averagi...              C   \n",
       " \n",
       "                                           model_answer  score  \\\n",
       " 0    To determine the fastest horse out of six, rac...    100   \n",
       " 1    ```json\\n{\\n  \"ANSWER\": \"D\"\\n}\\n```\\n\\nExplana...    100   \n",
       " 2    ```json\\n{\\n  \"ANSWER\": \"A\",\\n  \"explanation\":...    100   \n",
       " 3    To solve this problem, we need to plan the cro...      0   \n",
       " 4    {\\n  \"ANSWER\": \"C\",\\n  \"explanation\": \"Given t...      0   \n",
       " ..                                                 ...    ...   \n",
       " 295  ```json\\n{\"ANSWER\": \"B\"}\\n```\\n\\nExplanation: ...      0   \n",
       " 296  ```json\\n{\"ANSWER\": \"B\"}\\n```\\n\\nExplanation: ...      0   \n",
       " 297  The mass of the tree and the mass of the soil ...    100   \n",
       " 298  ```json\\n{\"ANSWER\": \"C\"}\\n```\\n\\nExplanation: ...      0   \n",
       " 299  To solve this problem, let's first understand ...    100   \n",
       " \n",
       "                                            json_answer json_answer_letter  \\\n",
       " 0                                      {'ANSWER': 'D'}                  D   \n",
       " 1                                      {'ANSWER': 'D'}                  D   \n",
       " 2    {'ANSWER': 'A', 'explanation': 'Initially, the...                  A   \n",
       " 3                                      {'ANSWER': 'A'}                  A   \n",
       " 4    {'ANSWER': 'C', 'explanation': 'Given the info...                  C   \n",
       " ..                                                 ...                ...   \n",
       " 295                                    {'ANSWER': 'B'}                  B   \n",
       " 296                                    {'ANSWER': 'B'}                  B   \n",
       " 297                                    {'ANSWER': 'B'}                  B   \n",
       " 298                                    {'ANSWER': 'C'}                  C   \n",
       " 299                                    {'ANSWER': 'C'}                  C   \n",
       " \n",
       "      invalid_answer_letter  \n",
       " 0                        0  \n",
       " 1                        0  \n",
       " 2                        0  \n",
       " 3                        0  \n",
       " 4                        0  \n",
       " ..                     ...  \n",
       " 295                      0  \n",
       " 296                      0  \n",
       " 297                      0  \n",
       " 298                      0  \n",
       " 299                      0  \n",
       " \n",
       " [300 rows x 14 columns],\n",
       " 'gpt-4o-mini-2024-07-18':      level_0  index         category  \\\n",
       " 0          0      1           Puzzle   \n",
       " 1          1      2           Puzzle   \n",
       " 2          2      3          Spatial   \n",
       " 3          3      4           Puzzle   \n",
       " 4          4      5           Puzzle   \n",
       " ..       ...    ...              ...   \n",
       " 295      295     26           Puzzle   \n",
       " 296      296     27           Puzzle   \n",
       " 297      297     28  Popular science   \n",
       " 298      298     29       Relational   \n",
       " 299      299     30  Popular science   \n",
       " \n",
       "                                               question  \\\n",
       " 0    You have six horses and want to race them to s...   \n",
       " 1    Suppose you're on a game show, and you're give...   \n",
       " 2    You are playing Russian roulette with a six-sh...   \n",
       " 3    A farmer wants to cross a river and take with ...   \n",
       " 4    Bob has three boxes in front of him - Box A, B...   \n",
       " ..                                                 ...   \n",
       " 295  A runaway trolley is heading down the tracks a...   \n",
       " 296  How do you efficiently measure exactly 4 gallo...   \n",
       " 297  A 2kg tree grows in a planted pot with 10kg of...   \n",
       " 298  A man steps over a 0.5m hurdle. What is the mi...   \n",
       " 299  I walk to my friend's house, averaging 3mph. H...   \n",
       " \n",
       "                                           human_answer  \\\n",
       " 0    Race them on a single race track with at least...   \n",
       " 1    It is not an advantage to switch. It makes no ...   \n",
       " 2    Yes, you should spin again. There was only one...   \n",
       " 3    Place the wolf, goat, and cabbage in separate ...   \n",
       " 4    Bob should not open a box because he has a 1/3...   \n",
       " ..                                                 ...   \n",
       " 295  It doesn't impact anyones lives, as the trolle...   \n",
       " 296                           Fill up the 4-gallon jug   \n",
       " 297  There is still 10kg of soil. Trees are primari...   \n",
       " 298  0m, he didn't jump at all, he stepped over the...   \n",
       " 299  Instantaneously fast, it is mathematically imp...   \n",
       " \n",
       "                              correct_answer  \\\n",
       " 0                                       One   \n",
       " 1    They have equal probability of winning   \n",
       " 2                Yes, you should spin again   \n",
       " 3                              One crossing   \n",
       " 4           No, Bob should not open any box   \n",
       " ..                                      ...   \n",
       " 295          It doesn't impact their lives.   \n",
       " 296              Another option not stated.   \n",
       " 297                                    10kg   \n",
       " 298                                      0m   \n",
       " 299                         Infinitely fast   \n",
       " \n",
       "                                        multiple_choice  \\\n",
       " 0                           [Two, Three, Unknown, One]   \n",
       " 1    [Door No.2, Door No.3, Door No.1, They have eq...   \n",
       " 2    [Yes, you should spin again, No, you should no...   \n",
       " 3    [Five crossings, The animals are not able to b...   \n",
       " 4    [Yes, Box A, No, Bob should not open any box, ...   \n",
       " ..                                                 ...   \n",
       " 295  [Pulling the lever will take their lives., Pul...   \n",
       " 296  [Another option not stated., Filling the 2-gal...   \n",
       " 297                             [9kg, 10kg, 11kg, 8kg]   \n",
       " 298  [0.5m, 0m, Slightly higher than 0.5m, Between ...   \n",
       " 299               [6mph, 9mph, Infinitely fast, 12mph]   \n",
       " \n",
       "                                  multi_choice_question correct_letter  \\\n",
       " 0    QUESTION\\nYou have six horses and want to race...              D   \n",
       " 1    QUESTION\\nSuppose you're on a game show, and y...              D   \n",
       " 2    QUESTION\\nYou are playing Russian roulette wit...              A   \n",
       " 3    QUESTION\\nA farmer wants to cross a river and ...              C   \n",
       " 4    QUESTION\\nBob has three boxes in front of him ...              B   \n",
       " ..                                                 ...            ...   \n",
       " 295  QUESTION\\nA runaway trolley is heading down th...              C   \n",
       " 296  QUESTION\\nHow do you efficiently measure exact...              A   \n",
       " 297  QUESTION\\nA 2kg tree grows in a planted pot wi...              B   \n",
       " 298  QUESTION\\nA man steps over a 0.5m hurdle. What...              B   \n",
       " 299  QUESTION\\nI walk to my friend's house, averagi...              C   \n",
       " \n",
       "                                           model_answer  score  \\\n",
       " 0    To determine the fastest horse among six, we n...      0   \n",
       " 1    The scenario presented is reminiscent of the M...      0   \n",
       " 2    To analyze the situation, let's break down the...    100   \n",
       " 3    To solve the problem, we need to ensure that t...      0   \n",
       " 4                                      {\"ANSWER\": \"C\"}      0   \n",
       " ..                                                 ...    ...   \n",
       " 295                                    {\"ANSWER\": \"D\"}      0   \n",
       " 296  To measure exactly 4 gallons of water using th...      0   \n",
       " 297  The weight of the soil in the pot does not cha...    100   \n",
       " 298  To determine the minimum height a man could ha...      0   \n",
       " 299  To find out how fast I need to run back to dou...    100   \n",
       " \n",
       "          json_answer json_answer_letter  invalid_answer_letter  \n",
       " 0    {'ANSWER': 'C'}                  C                      0  \n",
       " 1    {'ANSWER': 'A'}                  A                      0  \n",
       " 2    {'ANSWER': 'A'}                  A                      0  \n",
       " 3    {'ANSWER': 'D'}                  D                      0  \n",
       " 4    {'ANSWER': 'C'}                  C                      0  \n",
       " ..               ...                ...                    ...  \n",
       " 295  {'ANSWER': 'D'}                  D                      0  \n",
       " 296  {'ANSWER': 'D'}                  D                      0  \n",
       " 297  {'ANSWER': 'B'}                  B                      0  \n",
       " 298  {'ANSWER': 'C'}                  C                      0  \n",
       " 299  {'ANSWER': 'C'}                  C                      0  \n",
       " \n",
       " [300 rows x 14 columns],\n",
       " 'gpt-4o':      level_0  index         category  \\\n",
       " 0          0      1           Puzzle   \n",
       " 1          1      2           Puzzle   \n",
       " 2          2      3          Spatial   \n",
       " 3          3      4           Puzzle   \n",
       " 4          4      5           Puzzle   \n",
       " ..       ...    ...              ...   \n",
       " 295      295     26           Puzzle   \n",
       " 296      296     27           Puzzle   \n",
       " 297      297     28  Popular science   \n",
       " 298      298     29       Relational   \n",
       " 299      299     30  Popular science   \n",
       " \n",
       "                                               question  \\\n",
       " 0    You have six horses and want to race them to s...   \n",
       " 1    Suppose you're on a game show, and you're give...   \n",
       " 2    You are playing Russian roulette with a six-sh...   \n",
       " 3    A farmer wants to cross a river and take with ...   \n",
       " 4    Bob has three boxes in front of him - Box A, B...   \n",
       " ..                                                 ...   \n",
       " 295  A runaway trolley is heading down the tracks a...   \n",
       " 296  How do you efficiently measure exactly 4 gallo...   \n",
       " 297  A 2kg tree grows in a planted pot with 10kg of...   \n",
       " 298  A man steps over a 0.5m hurdle. What is the mi...   \n",
       " 299  I walk to my friend's house, averaging 3mph. H...   \n",
       " \n",
       "                                           human_answer  \\\n",
       " 0    Race them on a single race track with at least...   \n",
       " 1    It is not an advantage to switch. It makes no ...   \n",
       " 2    Yes, you should spin again. There was only one...   \n",
       " 3    Place the wolf, goat, and cabbage in separate ...   \n",
       " 4    Bob should not open a box because he has a 1/3...   \n",
       " ..                                                 ...   \n",
       " 295  It doesn't impact anyones lives, as the trolle...   \n",
       " 296                           Fill up the 4-gallon jug   \n",
       " 297  There is still 10kg of soil. Trees are primari...   \n",
       " 298  0m, he didn't jump at all, he stepped over the...   \n",
       " 299  Instantaneously fast, it is mathematically imp...   \n",
       " \n",
       "                              correct_answer  \\\n",
       " 0                                       One   \n",
       " 1    They have equal probability of winning   \n",
       " 2                Yes, you should spin again   \n",
       " 3                              One crossing   \n",
       " 4           No, Bob should not open any box   \n",
       " ..                                      ...   \n",
       " 295          It doesn't impact their lives.   \n",
       " 296              Another option not stated.   \n",
       " 297                                    10kg   \n",
       " 298                                      0m   \n",
       " 299                         Infinitely fast   \n",
       " \n",
       "                                        multiple_choice  \\\n",
       " 0                           [Two, Three, Unknown, One]   \n",
       " 1    [Door No.2, Door No.3, Door No.1, They have eq...   \n",
       " 2    [Yes, you should spin again, No, you should no...   \n",
       " 3    [Five crossings, The animals are not able to b...   \n",
       " 4    [Yes, Box A, No, Bob should not open any box, ...   \n",
       " ..                                                 ...   \n",
       " 295  [Pulling the lever will take their lives., Pul...   \n",
       " 296  [Another option not stated., Filling the 2-gal...   \n",
       " 297                             [9kg, 10kg, 11kg, 8kg]   \n",
       " 298  [0.5m, 0m, Slightly higher than 0.5m, Between ...   \n",
       " 299               [6mph, 9mph, Infinitely fast, 12mph]   \n",
       " \n",
       "                                  multi_choice_question correct_letter  \\\n",
       " 0    QUESTION\\nYou have six horses and want to race...              D   \n",
       " 1    QUESTION\\nSuppose you're on a game show, and y...              D   \n",
       " 2    QUESTION\\nYou are playing Russian roulette wit...              A   \n",
       " 3    QUESTION\\nA farmer wants to cross a river and ...              C   \n",
       " 4    QUESTION\\nBob has three boxes in front of him ...              B   \n",
       " ..                                                 ...            ...   \n",
       " 295  QUESTION\\nA runaway trolley is heading down th...              C   \n",
       " 296  QUESTION\\nHow do you efficiently measure exact...              A   \n",
       " 297  QUESTION\\nA 2kg tree grows in a planted pot wi...              B   \n",
       " 298  QUESTION\\nA man steps over a 0.5m hurdle. What...              B   \n",
       " 299  QUESTION\\nI walk to my friend's house, averagi...              C   \n",
       " \n",
       "                                           model_answer  score  \\\n",
       " 0    To determine the fastest horse out of six, you...    100   \n",
       " 1    This problem is a variation of the well-known ...      0   \n",
       " 2    To determine whether you should spin the chamb...      0   \n",
       " 3    To solve this problem, we need to ensure that ...      0   \n",
       " 4    Bob should avoid opening Box A because it will...      0   \n",
       " ..                                                 ...    ...   \n",
       " 295  Explanation:\\nThe scenario described is a clas...      0   \n",
       " 296  To measure exactly 4 gallons of water using th...    100   \n",
       " 297  To determine how much soil is left when the tr...    100   \n",
       " 298  To determine the minimum height the man could ...      0   \n",
       " 299  To determine how fast you would need to run ba...    100   \n",
       " \n",
       "          json_answer json_answer_letter  invalid_answer_letter  \n",
       " 0    {'ANSWER': 'D'}                  D                      0  \n",
       " 1    {'ANSWER': 'A'}                  A                      0  \n",
       " 2    {'ANSWER': 'B'}                  B                      0  \n",
       " 3    {'ANSWER': 'D'}                  D                      0  \n",
       " 4    {'ANSWER': 'C'}                  C                      0  \n",
       " ..               ...                ...                    ...  \n",
       " 295  {'ANSWER': 'B'}                  B                      0  \n",
       " 296  {'ANSWER': 'A'}                  A                      0  \n",
       " 297  {'ANSWER': 'B'}                  B                      0  \n",
       " 298  {'ANSWER': 'C'}                  C                      0  \n",
       " 299  {'ANSWER': 'C'}                  C                      0  \n",
       " \n",
       " [300 rows x 14 columns],\n",
       " 'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com':      level_0  index         category  \\\n",
       " 0          0      1           Puzzle   \n",
       " 1          1      2           Puzzle   \n",
       " 2          2      3          Spatial   \n",
       " 3          3      4           Puzzle   \n",
       " 4          4      5           Puzzle   \n",
       " ..       ...    ...              ...   \n",
       " 295      295     26           Puzzle   \n",
       " 296      296     27           Puzzle   \n",
       " 297      297     28  Popular science   \n",
       " 298      298     29       Relational   \n",
       " 299      299     30  Popular science   \n",
       " \n",
       "                                               question  \\\n",
       " 0    You have six horses and want to race them to s...   \n",
       " 1    Suppose you're on a game show, and you're give...   \n",
       " 2    You are playing Russian roulette with a six-sh...   \n",
       " 3    A farmer wants to cross a river and take with ...   \n",
       " 4    Bob has three boxes in front of him - Box A, B...   \n",
       " ..                                                 ...   \n",
       " 295  A runaway trolley is heading down the tracks a...   \n",
       " 296  How do you efficiently measure exactly 4 gallo...   \n",
       " 297  A 2kg tree grows in a planted pot with 10kg of...   \n",
       " 298  A man steps over a 0.5m hurdle. What is the mi...   \n",
       " 299  I walk to my friend's house, averaging 3mph. H...   \n",
       " \n",
       "                                           human_answer  \\\n",
       " 0    Race them on a single race track with at least...   \n",
       " 1    It is not an advantage to switch. It makes no ...   \n",
       " 2    Yes, you should spin again. There was only one...   \n",
       " 3    Place the wolf, goat, and cabbage in separate ...   \n",
       " 4    Bob should not open a box because he has a 1/3...   \n",
       " ..                                                 ...   \n",
       " 295  It doesn't impact anyones lives, as the trolle...   \n",
       " 296                           Fill up the 4-gallon jug   \n",
       " 297  There is still 10kg of soil. Trees are primari...   \n",
       " 298  0m, he didn't jump at all, he stepped over the...   \n",
       " 299  Instantaneously fast, it is mathematically imp...   \n",
       " \n",
       "                              correct_answer  \\\n",
       " 0                                       One   \n",
       " 1    They have equal probability of winning   \n",
       " 2                Yes, you should spin again   \n",
       " 3                              One crossing   \n",
       " 4           No, Bob should not open any box   \n",
       " ..                                      ...   \n",
       " 295          It doesn't impact their lives.   \n",
       " 296              Another option not stated.   \n",
       " 297                                    10kg   \n",
       " 298                                      0m   \n",
       " 299                         Infinitely fast   \n",
       " \n",
       "                                        multiple_choice  \\\n",
       " 0                           [One, Three, Unknown, Two]   \n",
       " 1    [Door No.2, Door No.1, Door No.3, They have eq...   \n",
       " 2    [No, you should not spin again, The probabilit...   \n",
       " 3    [Five crossings, The animals are not able to b...   \n",
       " 4    [Yes, Box C, No, Bob should not open any box, ...   \n",
       " ..                                                 ...   \n",
       " 295  [Pulling the lever will save their lives., Pul...   \n",
       " 296  [Filling the 2-gallon jug, twice., Filling the...   \n",
       " 297                             [10kg, 9kg, 8kg, 11kg]   \n",
       " 298  [Between 0.1m and 0.5m, 0m, 0.5m, Slightly hig...   \n",
       " 299               [12mph, 6mph, Infinitely fast, 9mph]   \n",
       " \n",
       "                                  multi_choice_question correct_letter  \\\n",
       " 0    QUESTION\\nYou have six horses and want to race...              A   \n",
       " 1    QUESTION\\nSuppose you're on a game show, and y...              D   \n",
       " 2    QUESTION\\nYou are playing Russian roulette wit...              D   \n",
       " 3    QUESTION\\nA farmer wants to cross a river and ...              D   \n",
       " 4    QUESTION\\nBob has three boxes in front of him ...              B   \n",
       " ..                                                 ...            ...   \n",
       " 295  QUESTION\\nA runaway trolley is heading down th...              C   \n",
       " 296  QUESTION\\nHow do you efficiently measure exact...              C   \n",
       " 297  QUESTION\\nA 2kg tree grows in a planted pot wi...              A   \n",
       " 298  QUESTION\\nA man steps over a 0.5m hurdle. What...              B   \n",
       " 299  QUESTION\\nI walk to my friend's house, averagi...              C   \n",
       " \n",
       "                                           model_answer  score  \\\n",
       " 0    To determine the minimum number of races neede...      0   \n",
       " 1    This problem is a variation of the classic Mon...      0   \n",
       " 2    To approach this problem, let's break down the...    100   \n",
       " 3    To solve this problem, we need to ensure that ...      0   \n",
       " 4    To solve this, let's consider the information ...    100   \n",
       " ..                                                 ...    ...   \n",
       " 295  To address this question, let's break down the...    100   \n",
       " 296  To measure exactly 4 gallons of water efficien...      0   \n",
       " 297  To solve this, let's analyze the situation:\\n\\...      0   \n",
       " 298  To determine the minimum height the man could ...      0   \n",
       " 299  To solve this problem, let's break it down ste...    100   \n",
       " \n",
       "          json_answer json_answer_letter  invalid_answer_letter  \n",
       " 0    {'ANSWER': 'B'}                  B                      0  \n",
       " 1    {'ANSWER': 'A'}                  A                      0  \n",
       " 2    {'ANSWER': 'D'}                  D                      0  \n",
       " 3    {'ANSWER': 'A'}                  A                      0  \n",
       " 4    {'ANSWER': 'B'}                  B                      0  \n",
       " ..               ...                ...                    ...  \n",
       " 295  {'ANSWER': 'C'}                  C                      0  \n",
       " 296  {'ANSWER': 'D'}                  D                      0  \n",
       " 297  {'ANSWER': 'B'}                  B                      0  \n",
       " 298  {'ANSWER': 'C'}                  C                      0  \n",
       " 299  {'ANSWER': 'C'}                  C                      0  \n",
       " \n",
       " [300 rows x 14 columns],\n",
       " 'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com':      level_0  index         category  \\\n",
       " 0          0      1           Puzzle   \n",
       " 1          1      2           Puzzle   \n",
       " 2          2      3          Spatial   \n",
       " 3          3      4           Puzzle   \n",
       " 4          4      5           Puzzle   \n",
       " ..       ...    ...              ...   \n",
       " 295      295     26           Puzzle   \n",
       " 296      296     27           Puzzle   \n",
       " 297      297     28  Popular science   \n",
       " 298      298     29       Relational   \n",
       " 299      299     30  Popular science   \n",
       " \n",
       "                                               question  \\\n",
       " 0    You have six horses and want to race them to s...   \n",
       " 1    Suppose you're on a game show, and you're give...   \n",
       " 2    You are playing Russian roulette with a six-sh...   \n",
       " 3    A farmer wants to cross a river and take with ...   \n",
       " 4    Bob has three boxes in front of him - Box A, B...   \n",
       " ..                                                 ...   \n",
       " 295  A runaway trolley is heading down the tracks a...   \n",
       " 296  How do you efficiently measure exactly 4 gallo...   \n",
       " 297  A 2kg tree grows in a planted pot with 10kg of...   \n",
       " 298  A man steps over a 0.5m hurdle. What is the mi...   \n",
       " 299  I walk to my friend's house, averaging 3mph. H...   \n",
       " \n",
       "                                           human_answer  \\\n",
       " 0    Race them on a single race track with at least...   \n",
       " 1    It is not an advantage to switch. It makes no ...   \n",
       " 2    Yes, you should spin again. There was only one...   \n",
       " 3    Place the wolf, goat, and cabbage in separate ...   \n",
       " 4    Bob should not open a box because he has a 1/3...   \n",
       " ..                                                 ...   \n",
       " 295  It doesn't impact anyones lives, as the trolle...   \n",
       " 296                           Fill up the 4-gallon jug   \n",
       " 297  There is still 10kg of soil. Trees are primari...   \n",
       " 298  0m, he didn't jump at all, he stepped over the...   \n",
       " 299  Instantaneously fast, it is mathematically imp...   \n",
       " \n",
       "                              correct_answer  \\\n",
       " 0                                       One   \n",
       " 1    They have equal probability of winning   \n",
       " 2                Yes, you should spin again   \n",
       " 3                              One crossing   \n",
       " 4           No, Bob should not open any box   \n",
       " ..                                      ...   \n",
       " 295          It doesn't impact their lives.   \n",
       " 296              Another option not stated.   \n",
       " 297                                    10kg   \n",
       " 298                                      0m   \n",
       " 299                         Infinitely fast   \n",
       " \n",
       "                                        multiple_choice  \\\n",
       " 0                           [Two, One, Three, Unknown]   \n",
       " 1    [Door No.2, They have equal probability of win...   \n",
       " 2    [Yes, you should spin again, It makes no diffe...   \n",
       " 3    [Five crossings, Seven crossings, The animals ...   \n",
       " 4    [Yes, Box A, Yes, Box B, No, Bob should not op...   \n",
       " ..                                                 ...   \n",
       " 295  [This is an ethical dilemma., It doesn't impac...   \n",
       " 296  [Another option not stated., Filling the 5-gal...   \n",
       " 297                             [10kg, 11kg, 8kg, 9kg]   \n",
       " 298  [0m, Slightly higher than 0.5m, Between 0.1m a...   \n",
       " 299               [9mph, Infinitely fast, 6mph, 12mph]   \n",
       " \n",
       "                                  multi_choice_question correct_letter  \\\n",
       " 0    QUESTION\\nYou have six horses and want to race...              B   \n",
       " 1    QUESTION\\nSuppose you're on a game show, and y...              B   \n",
       " 2    QUESTION\\nYou are playing Russian roulette wit...              A   \n",
       " 3    QUESTION\\nA farmer wants to cross a river and ...              D   \n",
       " 4    QUESTION\\nBob has three boxes in front of him ...              C   \n",
       " ..                                                 ...            ...   \n",
       " 295  QUESTION\\nA runaway trolley is heading down th...              B   \n",
       " 296  QUESTION\\nHow do you efficiently measure exact...              A   \n",
       " 297  QUESTION\\nA 2kg tree grows in a planted pot wi...              A   \n",
       " 298  QUESTION\\nA man steps over a 0.5m hurdle. What...              A   \n",
       " 299  QUESTION\\nI walk to my friend's house, averagi...              B   \n",
       " \n",
       "                                           model_answer  score  \\\n",
       " 0    To determine the minimum number of races neede...    100   \n",
       " 1    Here's the explanation and answer:\\n\\n**Explan...      0   \n",
       " 2    {\"ANSWER\": \"B\"}\\n\\nExplanation:\\nWhen your opp...      0   \n",
       " 3    To solve this classic puzzle, let's break it d...      0   \n",
       " 4    To determine the best course of action for Bob...      0   \n",
       " ..                                                 ...    ...   \n",
       " 295  To answer this question, let's break down the ...      0   \n",
       " 296  To measure exactly 4 gallons of water with the...      0   \n",
       " 297  To solve this problem, let's consider the init...      0   \n",
       " 298  To determine the minimum height the man could ...      0   \n",
       " 299  To solve this problem, let's break it down ste...    100   \n",
       " \n",
       "          json_answer json_answer_letter  invalid_answer_letter  \n",
       " 0    {'ANSWER': 'B'}                  B                      0  \n",
       " 1    {'ANSWER': 'D'}                  D                      0  \n",
       " 2    {'ANSWER': 'B'}                  B                      0  \n",
       " 3    {'ANSWER': 'A'}                  A                      0  \n",
       " 4    {'ANSWER': 'B'}                  B                      0  \n",
       " ..               ...                ...                    ...  \n",
       " 295  {'ANSWER': 'D'}                  D                      0  \n",
       " 296  {'ANSWER': 'C'}                  C                      0  \n",
       " 297  {'ANSWER': 'D'}                  D                      0  \n",
       " 298  {'ANSWER': 'B'}                  B                      0  \n",
       " 299  {'ANSWER': 'B'}                  B                      0  \n",
       " \n",
       " [300 rows x 14 columns],\n",
       " 'meta_llama3-70b-instruct-v1_0':      level_0  index         category  \\\n",
       " 0          0      1           Puzzle   \n",
       " 1          1      2           Puzzle   \n",
       " 2          2      3          Spatial   \n",
       " 3          3      4           Puzzle   \n",
       " 4          4      5           Puzzle   \n",
       " ..       ...    ...              ...   \n",
       " 295      295     26           Puzzle   \n",
       " 296      296     27           Puzzle   \n",
       " 297      297     28  Popular science   \n",
       " 298      298     29       Relational   \n",
       " 299      299     30  Popular science   \n",
       " \n",
       "                                               question  \\\n",
       " 0    You have six horses and want to race them to s...   \n",
       " 1    Suppose you're on a game show, and you're give...   \n",
       " 2    You are playing Russian roulette with a six-sh...   \n",
       " 3    A farmer wants to cross a river and take with ...   \n",
       " 4    Bob has three boxes in front of him - Box A, B...   \n",
       " ..                                                 ...   \n",
       " 295  A runaway trolley is heading down the tracks a...   \n",
       " 296  How do you efficiently measure exactly 4 gallo...   \n",
       " 297  A 2kg tree grows in a planted pot with 10kg of...   \n",
       " 298  A man steps over a 0.5m hurdle. What is the mi...   \n",
       " 299  I walk to my friend's house, averaging 3mph. H...   \n",
       " \n",
       "                                           human_answer  \\\n",
       " 0    Race them on a single race track with at least...   \n",
       " 1    It is not an advantage to switch. It makes no ...   \n",
       " 2    Yes, you should spin again. There was only one...   \n",
       " 3    Place the wolf, goat, and cabbage in separate ...   \n",
       " 4    Bob should not open a box because he has a 1/3...   \n",
       " ..                                                 ...   \n",
       " 295  It doesn't impact anyones lives, as the trolle...   \n",
       " 296                           Fill up the 4-gallon jug   \n",
       " 297  There is still 10kg of soil. Trees are primari...   \n",
       " 298  0m, he didn't jump at all, he stepped over the...   \n",
       " 299  Instantaneously fast, it is mathematically imp...   \n",
       " \n",
       "                              correct_answer  \\\n",
       " 0                                       One   \n",
       " 1    They have equal probability of winning   \n",
       " 2                Yes, you should spin again   \n",
       " 3                              One crossing   \n",
       " 4           No, Bob should not open any box   \n",
       " ..                                      ...   \n",
       " 295          It doesn't impact their lives.   \n",
       " 296              Another option not stated.   \n",
       " 297                                    10kg   \n",
       " 298                                      0m   \n",
       " 299                         Infinitely fast   \n",
       " \n",
       "                                        multiple_choice  \\\n",
       " 0                           [Three, One, Two, Unknown]   \n",
       " 1    [Door No.3, Door No.1, Door No.2, They have eq...   \n",
       " 2    [The probabilities are random and unknown, It ...   \n",
       " 3    [The animals are not able to be transported ac...   \n",
       " 4    [Yes, Box A, Yes, Box C, Yes, Box B, No, Bob s...   \n",
       " ..                                                 ...   \n",
       " 295  [Pulling the lever will take their lives., Pul...   \n",
       " 296  [Another option not stated., Filling the 2-gal...   \n",
       " 297                             [9kg, 10kg, 11kg, 8kg]   \n",
       " 298  [0.5m, 0m, Slightly higher than 0.5m, Between ...   \n",
       " 299               [6mph, 9mph, Infinitely fast, 12mph]   \n",
       " \n",
       "                                  multi_choice_question correct_letter  \\\n",
       " 0    QUESTION\\nYou have six horses and want to race...              B   \n",
       " 1    QUESTION\\nSuppose you're on a game show, and y...              D   \n",
       " 2    QUESTION\\nYou are playing Russian roulette wit...              C   \n",
       " 3    QUESTION\\nA farmer wants to cross a river and ...              D   \n",
       " 4    QUESTION\\nBob has three boxes in front of him ...              D   \n",
       " ..                                                 ...            ...   \n",
       " 295  QUESTION\\nA runaway trolley is heading down th...              C   \n",
       " 296  QUESTION\\nHow do you efficiently measure exact...              A   \n",
       " 297  QUESTION\\nA 2kg tree grows in a planted pot wi...              B   \n",
       " 298  QUESTION\\nA man steps over a 0.5m hurdle. What...              B   \n",
       " 299  QUESTION\\nI walk to my friend's house, averagi...              C   \n",
       " \n",
       "                                           model_answer  score  \\\n",
       " 0    Here's my explanation and answer:\\n\\nTo find t...      0   \n",
       " 1    Here's my explanation and answer:\\n\\nWhen you ...      0   \n",
       " 2    Here's my explanation and answer:\\n\\nWhen the ...    100   \n",
       " 3    Here's my explanation and answer:\\n\\nTo solve ...      0   \n",
       " 4    Here's my explanation and answer:\\n\\nSince Bob...      0   \n",
       " ..                                                 ...    ...   \n",
       " 295  Here's my explanation and answer:\\n\\nThe class...      0   \n",
       " 296  Here's my explanation and answer:\\n\\nTo measur...      0   \n",
       " 297  Here's my explanation and answer:\\n\\nThe quest...      0   \n",
       " 298  Here's my explanation and answer:\\n\\nTo clear ...      0   \n",
       " 299  Here's my explanation and answer:\\n\\nLet's bre...    100   \n",
       " \n",
       "          json_answer json_answer_letter  invalid_answer_letter  \n",
       " 0    {'ANSWER': 'A'}                  A                      0  \n",
       " 1    {'ANSWER': 'C'}                  C                      0  \n",
       " 2    {'ANSWER': 'C'}                  C                      0  \n",
       " 3    {'ANSWER': 'C'}                  C                      0  \n",
       " 4    {'ANSWER': 'C'}                  C                      0  \n",
       " ..               ...                ...                    ...  \n",
       " 295  {'ANSWER': 'B'}                  B                      0  \n",
       " 296  {'ANSWER': 'D'}                  D                      0  \n",
       " 297  {'ANSWER': 'A'}                  A                      0  \n",
       " 298  {'ANSWER': 'C'}                  C                      0  \n",
       " 299  {'ANSWER': 'C'}                  C                      0  \n",
       " \n",
       " [300 rows x 14 columns],\n",
       " 'mistral-large-latest':      level_0  index         category  \\\n",
       " 0          0      1           Puzzle   \n",
       " 1          1      2           Puzzle   \n",
       " 2          2      3          Spatial   \n",
       " 3          3      4           Puzzle   \n",
       " 4          4      5           Puzzle   \n",
       " ..       ...    ...              ...   \n",
       " 295      295     26           Puzzle   \n",
       " 296      296     27           Puzzle   \n",
       " 297      297     28  Popular science   \n",
       " 298      298     29       Relational   \n",
       " 299      299     30  Popular science   \n",
       " \n",
       "                                               question  \\\n",
       " 0    You have six horses and want to race them to s...   \n",
       " 1    Suppose you're on a game show, and you're give...   \n",
       " 2    You are playing Russian roulette with a six-sh...   \n",
       " 3    A farmer wants to cross a river and take with ...   \n",
       " 4    Bob has three boxes in front of him - Box A, B...   \n",
       " ..                                                 ...   \n",
       " 295  A runaway trolley is heading down the tracks a...   \n",
       " 296  How do you efficiently measure exactly 4 gallo...   \n",
       " 297  A 2kg tree grows in a planted pot with 10kg of...   \n",
       " 298  A man steps over a 0.5m hurdle. What is the mi...   \n",
       " 299  I walk to my friend's house, averaging 3mph. H...   \n",
       " \n",
       "                                           human_answer  \\\n",
       " 0    Race them on a single race track with at least...   \n",
       " 1    It is not an advantage to switch. It makes no ...   \n",
       " 2    Yes, you should spin again. There was only one...   \n",
       " 3    Place the wolf, goat, and cabbage in separate ...   \n",
       " 4    Bob should not open a box because he has a 1/3...   \n",
       " ..                                                 ...   \n",
       " 295  It doesn't impact anyones lives, as the trolle...   \n",
       " 296                           Fill up the 4-gallon jug   \n",
       " 297  There is still 10kg of soil. Trees are primari...   \n",
       " 298  0m, he didn't jump at all, he stepped over the...   \n",
       " 299  Instantaneously fast, it is mathematically imp...   \n",
       " \n",
       "                              correct_answer  \\\n",
       " 0                                       One   \n",
       " 1    They have equal probability of winning   \n",
       " 2                Yes, you should spin again   \n",
       " 3                              One crossing   \n",
       " 4           No, Bob should not open any box   \n",
       " ..                                      ...   \n",
       " 295          It doesn't impact their lives.   \n",
       " 296              Another option not stated.   \n",
       " 297                                    10kg   \n",
       " 298                                      0m   \n",
       " 299                         Infinitely fast   \n",
       " \n",
       "                                        multiple_choice  \\\n",
       " 0                           [Two, One, Three, Unknown]   \n",
       " 1    [Door No.1, They have equal probability of win...   \n",
       " 2    [Yes, you should spin again, No, you should no...   \n",
       " 3    [One crossing, The animals are not able to be ...   \n",
       " 4    [Yes, Box A, No, Bob should not open any box, ...   \n",
       " ..                                                 ...   \n",
       " 295  [It doesn't impact their lives., Pulling the l...   \n",
       " 296  [Filling the 2-gallon jug, twice., Another opt...   \n",
       " 297                             [9kg, 10kg, 11kg, 8kg]   \n",
       " 298  [Slightly higher than 0.5m, 0.5m, 0m, Between ...   \n",
       " 299               [Infinitely fast, 12mph, 9mph, 6mph]   \n",
       " \n",
       "                                  multi_choice_question correct_letter  \\\n",
       " 0    QUESTION\\nYou have six horses and want to race...              B   \n",
       " 1    QUESTION\\nSuppose you're on a game show, and y...              B   \n",
       " 2    QUESTION\\nYou are playing Russian roulette wit...              A   \n",
       " 3    QUESTION\\nA farmer wants to cross a river and ...              A   \n",
       " 4    QUESTION\\nBob has three boxes in front of him ...              B   \n",
       " ..                                                 ...            ...   \n",
       " 295  QUESTION\\nA runaway trolley is heading down th...              A   \n",
       " 296  QUESTION\\nHow do you efficiently measure exact...              B   \n",
       " 297  QUESTION\\nA 2kg tree grows in a planted pot wi...              B   \n",
       " 298  QUESTION\\nA man steps over a 0.5m hurdle. What...              C   \n",
       " 299  QUESTION\\nI walk to my friend's house, averagi...              A   \n",
       " \n",
       "                                           model_answer  score  \\\n",
       " 0    ```json\\n{\\n  \"ANSWER\": \"C\",\\n  \"EXPLANATION\":...      0   \n",
       " 1    ```json\\n{\\n  \"ANSWER\": \"D\",\\n  \"EXPLANATION\":...      0   \n",
       " 2    ```json\\n{\\n  \"ANSWER\": \"B\",\\n  \"EXPLANATION\":...      0   \n",
       " 3    ```json\\n{\\n  \"ANSWER\": \"D\",\\n  \"EXPLANATION\":...      0   \n",
       " 4    ```json\\n{\\n  \"ANSWER\": \"C\",\\n  \"EXPLANATION\":...      0   \n",
       " ..                                                 ...    ...   \n",
       " 295  ```json\\n{\\n  \"ANSWER\": \"C\",\\n  \"EXPLANATION\":...      0   \n",
       " 296  ```json\\n{\\n  \"ANSWER\": \"D\",\\n  \"EXPLANATION\":...      0   \n",
       " 297  ```json\\n{\\n  \"ANSWER\": \"B\"\\n}\\n```\\n\\nExplana...    100   \n",
       " 298  ```json\\n{\\n  \"ANSWER\": \"C\",\\n  \"EXPLANATION\":...    100   \n",
       " 299  To solve this problem, let's break it down ste...    100   \n",
       " \n",
       "                                            json_answer json_answer_letter  \\\n",
       " 0    {'ANSWER': 'C', 'EXPLANATION': 'To determine t...                  C   \n",
       " 1    {'ANSWER': 'D', 'EXPLANATION': \"This is a vari...                  D   \n",
       " 2    {'ANSWER': 'B', 'EXPLANATION': 'When your oppo...                  B   \n",
       " 3    {'ANSWER': 'D', 'EXPLANATION': 'The farmer can...                  D   \n",
       " 4    {'ANSWER': 'C', 'EXPLANATION': 'Bob should ope...                  C   \n",
       " ..                                                 ...                ...   \n",
       " 295  {'ANSWER': 'C', 'EXPLANATION': 'Pulling the le...                  C   \n",
       " 296  {'ANSWER': 'D', 'EXPLANATION': 'To measure exa...                  D   \n",
       " 297                                    {'ANSWER': 'B'}                  B   \n",
       " 298  {'ANSWER': 'C', 'EXPLANATION': 'The man could ...                  C   \n",
       " 299                                    {'ANSWER': 'A'}                  A   \n",
       " \n",
       "      invalid_answer_letter  \n",
       " 0                        0  \n",
       " 1                        0  \n",
       " 2                        0  \n",
       " 3                        0  \n",
       " 4                        0  \n",
       " ..                     ...  \n",
       " 295                      0  \n",
       " 296                      0  \n",
       " 297                      0  \n",
       " 298                      0  \n",
       " 299                      0  \n",
       " \n",
       " [300 rows x 14 columns],\n",
       " 'open-mixtral-8x22b':      level_0  index         category  \\\n",
       " 0          0      1           Puzzle   \n",
       " 1          1      2           Puzzle   \n",
       " 2          2      3          Spatial   \n",
       " 3          3      4           Puzzle   \n",
       " 4          4      5           Puzzle   \n",
       " ..       ...    ...              ...   \n",
       " 295      295     26           Puzzle   \n",
       " 296      296     27           Puzzle   \n",
       " 297      297     28  Popular science   \n",
       " 298      298     29       Relational   \n",
       " 299      299     30  Popular science   \n",
       " \n",
       "                                               question  \\\n",
       " 0    You have six horses and want to race them to s...   \n",
       " 1    Suppose you're on a game show, and you're give...   \n",
       " 2    You are playing Russian roulette with a six-sh...   \n",
       " 3    A farmer wants to cross a river and take with ...   \n",
       " 4    Bob has three boxes in front of him - Box A, B...   \n",
       " ..                                                 ...   \n",
       " 295  A runaway trolley is heading down the tracks a...   \n",
       " 296  How do you efficiently measure exactly 4 gallo...   \n",
       " 297  A 2kg tree grows in a planted pot with 10kg of...   \n",
       " 298  A man steps over a 0.5m hurdle. What is the mi...   \n",
       " 299  I walk to my friend's house, averaging 3mph. H...   \n",
       " \n",
       "                                           human_answer  \\\n",
       " 0    Race them on a single race track with at least...   \n",
       " 1    It is not an advantage to switch. It makes no ...   \n",
       " 2    Yes, you should spin again. There was only one...   \n",
       " 3    Place the wolf, goat, and cabbage in separate ...   \n",
       " 4    Bob should not open a box because he has a 1/3...   \n",
       " ..                                                 ...   \n",
       " 295  It doesn't impact anyones lives, as the trolle...   \n",
       " 296                           Fill up the 4-gallon jug   \n",
       " 297  There is still 10kg of soil. Trees are primari...   \n",
       " 298  0m, he didn't jump at all, he stepped over the...   \n",
       " 299  Instantaneously fast, it is mathematically imp...   \n",
       " \n",
       "                              correct_answer  \\\n",
       " 0                                       One   \n",
       " 1    They have equal probability of winning   \n",
       " 2                Yes, you should spin again   \n",
       " 3                              One crossing   \n",
       " 4           No, Bob should not open any box   \n",
       " ..                                      ...   \n",
       " 295          It doesn't impact their lives.   \n",
       " 296              Another option not stated.   \n",
       " 297                                    10kg   \n",
       " 298                                      0m   \n",
       " 299                         Infinitely fast   \n",
       " \n",
       "                                        multiple_choice  \\\n",
       " 0                           [Three, One, Two, Unknown]   \n",
       " 1    [Door No.3, Door No.1, Door No.2, They have eq...   \n",
       " 2    [The probabilities are random and unknown, It ...   \n",
       " 3    [The animals are not able to be transported ac...   \n",
       " 4    [Yes, Box A, Yes, Box C, Yes, Box B, No, Bob s...   \n",
       " ..                                                 ...   \n",
       " 295  [Pulling the lever will take their lives., Pul...   \n",
       " 296  [Another option not stated., Filling the 2-gal...   \n",
       " 297                             [9kg, 10kg, 11kg, 8kg]   \n",
       " 298  [0.5m, 0m, Slightly higher than 0.5m, Between ...   \n",
       " 299               [6mph, 9mph, Infinitely fast, 12mph]   \n",
       " \n",
       "                                  multi_choice_question correct_letter  \\\n",
       " 0    QUESTION\\nYou have six horses and want to race...              B   \n",
       " 1    QUESTION\\nSuppose you're on a game show, and y...              D   \n",
       " 2    QUESTION\\nYou are playing Russian roulette wit...              C   \n",
       " 3    QUESTION\\nA farmer wants to cross a river and ...              D   \n",
       " 4    QUESTION\\nBob has three boxes in front of him ...              D   \n",
       " ..                                                 ...            ...   \n",
       " 295  QUESTION\\nA runaway trolley is heading down th...              C   \n",
       " 296  QUESTION\\nHow do you efficiently measure exact...              A   \n",
       " 297  QUESTION\\nA 2kg tree grows in a planted pot wi...              B   \n",
       " 298  QUESTION\\nA man steps over a 0.5m hurdle. What...              B   \n",
       " 299  QUESTION\\nI walk to my friend's house, averagi...              C   \n",
       " \n",
       "                                           model_answer  score  \\\n",
       " 0    To determine the fastest horse, you can use a ...      0   \n",
       " 1    The situation you've described is known as the...      0   \n",
       " 2    The correct answer is D. No, you should not sp...      0   \n",
       " 3    The minimum number of crossings the farmer nee...      0   \n",
       " 4    {\"ANSWER\": \"D\"}\\n\\nExplanation:\\nBob does not ...    100   \n",
       " ..                                                 ...    ...   \n",
       " 295  {\"ANSWER\": \"B\", \"EXPLANATION\": \"Pulling the le...      0   \n",
       " 296  To measure exactly 4 gallons of water with the...    100   \n",
       " 297  The weight of the soil does not change as the ...    100   \n",
       " 298  The minimum height a man could have jumped to ...      0   \n",
       " 299                                                         0   \n",
       " \n",
       "                                            json_answer json_answer_letter  \\\n",
       " 0                                      {'ANSWER': 'A'}                  A   \n",
       " 1                                      {'ANSWER': 'A'}                  A   \n",
       " 2                                      {'ANSWER': 'D'}                  D   \n",
       " 3                                      {'ANSWER': 'B'}                  B   \n",
       " 4                                      {'ANSWER': 'D'}                  D   \n",
       " ..                                                 ...                ...   \n",
       " 295  {'ANSWER': 'B', 'EXPLANATION': 'Pulling the le...                  B   \n",
       " 296                                    {'ANSWER': 'A'}                  A   \n",
       " 297                                    {'ANSWER': 'B'}                  B   \n",
       " 298                                    {'ANSWER': 'A'}                  A   \n",
       " 299                                               None               None   \n",
       " \n",
       "      invalid_answer_letter  \n",
       " 0                        0  \n",
       " 1                        0  \n",
       " 2                        0  \n",
       " 3                        0  \n",
       " 4                        0  \n",
       " ..                     ...  \n",
       " 295                      0  \n",
       " 296                      0  \n",
       " 297                      0  \n",
       " 298                      0  \n",
       " 299                      1  \n",
       " \n",
       " [300 rows x 14 columns]}"
      ]
     },
     "execution_count": 20,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "answers_save_path = './2024-07-20-Multi-Benchmark/auto_eval_outputs'\n",
    "\n",
    "all_llm_answers = load_all_llm_answers_from_json(answers_save_path, prefix_replace='auto_eval-')\n",
    "\n",
    "all_llm_answers"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "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>level_0</th>\n",
       "      <th>index</th>\n",
       "      <th>category</th>\n",
       "      <th>question</th>\n",
       "      <th>human_answer</th>\n",
       "      <th>correct_answer</th>\n",
       "      <th>multiple_choice</th>\n",
       "      <th>multi_choice_question</th>\n",
       "      <th>correct_letter</th>\n",
       "      <th>model_answer</th>\n",
       "      <th>score</th>\n",
       "      <th>json_answer</th>\n",
       "      <th>json_answer_letter</th>\n",
       "      <th>invalid_answer_letter</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>Puzzle</td>\n",
       "      <td>You have six horses and want to race them to s...</td>\n",
       "      <td>Race them on a single race track with at least...</td>\n",
       "      <td>One</td>\n",
       "      <td>[Three, One, Two, Unknown]</td>\n",
       "      <td>QUESTION\\nYou have six horses and want to race...</td>\n",
       "      <td>B</td>\n",
       "      <td>Let's think through this step-by-step:\\n\\n1. W...</td>\n",
       "      <td>0</td>\n",
       "      <td>{'ANSWER': 'C'}</td>\n",
       "      <td>C</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   level_0  index category                                           question  \\\n",
       "0        0      1   Puzzle  You have six horses and want to race them to s...   \n",
       "\n",
       "                                        human_answer correct_answer  \\\n",
       "0  Race them on a single race track with at least...            One   \n",
       "\n",
       "              multiple_choice  \\\n",
       "0  [Three, One, Two, Unknown]   \n",
       "\n",
       "                               multi_choice_question correct_letter  \\\n",
       "0  QUESTION\\nYou have six horses and want to race...              B   \n",
       "\n",
       "                                        model_answer  score      json_answer  \\\n",
       "0  Let's think through this step-by-step:\\n\\n1. W...      0  {'ANSWER': 'C'}   \n",
       "\n",
       "  json_answer_letter  invalid_answer_letter  \n",
       "0                  C                      0  "
      ]
     },
     "execution_count": 21,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "all_llm_answers[list(all_llm_answers.keys())[0]].head(1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [],
   "source": [
    "# # Sanity check to see if the scores are being calculated correctly in the below cells after a low of manipulation\n",
    "# model_scores = []\n",
    "# for llm_model in list(all_llm_answers.keys()):\n",
    "#     correct_answer = all_llm_answers[llm_model]['correct_letter'] == all_llm_answers[llm_model]['json_answer_letter']\n",
    "#     percentage_correct = correct_answer.value_counts(normalize=True)[True] * 100\n",
    "#     initial_score = all_llm_answers[llm_model]['score'].mean()\n",
    "#     dict_results = {\n",
    "#         'model': llm_model,\n",
    "#         'percentage_correct': percentage_correct,\n",
    "#         'initial_score': initial_score\n",
    "#     }\n",
    "#     model_scores.append(dict_results)\n",
    "\n",
    "# model_scores_series = pd.DataFrame(model_scores).sort_values(by='percentage_correct', ascending=False)\n",
    "# model_scores_series"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "metadata": {},
   "outputs": [],
   "source": [
    "moe_llm_answers = {}\n",
    "for llm_model in list(all_llm_answers.keys()):\n",
    "    answer_subset = all_llm_answers[llm_model]\n",
    "    data = answer_subset[[\n",
    "        'index', 'question', 'multiple_choice', 'correct_answer',\n",
    "        'multi_choice_question', 'correct_letter', 'json_answer_letter',\n",
    "    ]]\n",
    "    moe_llm_answers[llm_model] = []\n",
    "    for idx, row in data.iterrows():\n",
    "        # if idx > 0:\n",
    "        #     continue\n",
    "        normalized_choices_dict_inv = dict(zip(sorted(row['multiple_choice']), ['A', 'B', 'C', 'D']))\n",
    "        presented_choices_dict = dict(zip(['A', 'B', 'C', 'D'], row['multiple_choice']))\n",
    "        correct_answer = presented_choices_dict[row['correct_letter']]\n",
    "        row['correct'] = row['correct_letter'] == row['json_answer_letter']\n",
    "        assert(row['correct_answer'] == correct_answer)\n",
    "        normalized_correct_letter = normalized_choices_dict_inv[correct_answer]\n",
    "        if row['json_answer_letter'] not in presented_choices_dict:\n",
    "            selected_answer, normalized_choice_letter = None, None\n",
    "        else:\n",
    "            selected_answer = presented_choices_dict[row['json_answer_letter']]\n",
    "            normalized_choice_letter = normalized_choices_dict_inv[selected_answer]\n",
    "        moe_llm_answers[llm_model].append({\n",
    "            'question': row['question'],\n",
    "            # 'selected_answer': selected_answer,\n",
    "            # 'correct_answer': correct_answer,\n",
    "            'correct': row['correct'],\n",
    "            'normalized_correct_letter': normalized_correct_letter,\n",
    "            'normalized_choice_letter': normalized_choice_letter,\n",
    "        })\n",
    "        #print(f\"idx: {idx} | {llm_model} | {moe_llm_answers[llm_model]}\\n\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "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 tr th {\n",
       "        text-align: left;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr>\n",
       "      <th></th>\n",
       "      <th colspan=\"2\" halign=\"left\">bench</th>\n",
       "      <th>claude-3-5-sonnet-20240620</th>\n",
       "      <th>bench</th>\n",
       "      <th>claude-3-5-sonnet-20240620</th>\n",
       "      <th colspan=\"2\" halign=\"left\">claude-3-opus-20240229</th>\n",
       "      <th colspan=\"2\" halign=\"left\">command-r-plus</th>\n",
       "      <th>gemini-1_5-pro</th>\n",
       "      <th>...</th>\n",
       "      <th colspan=\"2\" halign=\"left\">Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com</th>\n",
       "      <th colspan=\"2\" halign=\"left\">Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com</th>\n",
       "      <th colspan=\"2\" halign=\"left\">meta_llama3-70b-instruct-v1_0</th>\n",
       "      <th colspan=\"2\" halign=\"left\">mistral-large-latest</th>\n",
       "      <th colspan=\"2\" halign=\"left\">open-mixtral-8x22b</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th></th>\n",
       "      <th>index</th>\n",
       "      <th>question</th>\n",
       "      <th>correct</th>\n",
       "      <th>normalized_correct_letter</th>\n",
       "      <th>normalized_choice_letter</th>\n",
       "      <th>correct</th>\n",
       "      <th>normalized_choice_letter</th>\n",
       "      <th>correct</th>\n",
       "      <th>normalized_choice_letter</th>\n",
       "      <th>correct</th>\n",
       "      <th>...</th>\n",
       "      <th>correct</th>\n",
       "      <th>normalized_choice_letter</th>\n",
       "      <th>correct</th>\n",
       "      <th>normalized_choice_letter</th>\n",
       "      <th>correct</th>\n",
       "      <th>normalized_choice_letter</th>\n",
       "      <th>correct</th>\n",
       "      <th>normalized_choice_letter</th>\n",
       "      <th>correct</th>\n",
       "      <th>normalized_choice_letter</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0</td>\n",
       "      <td>You have six horses and want to race them to s...</td>\n",
       "      <td>False</td>\n",
       "      <td>A</td>\n",
       "      <td>C</td>\n",
       "      <td>False</td>\n",
       "      <td>B</td>\n",
       "      <td>True</td>\n",
       "      <td>A</td>\n",
       "      <td>False</td>\n",
       "      <td>...</td>\n",
       "      <td>False</td>\n",
       "      <td>B</td>\n",
       "      <td>True</td>\n",
       "      <td>A</td>\n",
       "      <td>False</td>\n",
       "      <td>B</td>\n",
       "      <td>False</td>\n",
       "      <td>B</td>\n",
       "      <td>False</td>\n",
       "      <td>B</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1</td>\n",
       "      <td>Suppose you're on a game show, and you're give...</td>\n",
       "      <td>True</td>\n",
       "      <td>D</td>\n",
       "      <td>D</td>\n",
       "      <td>False</td>\n",
       "      <td>C</td>\n",
       "      <td>False</td>\n",
       "      <td>B</td>\n",
       "      <td>False</td>\n",
       "      <td>...</td>\n",
       "      <td>False</td>\n",
       "      <td>B</td>\n",
       "      <td>False</td>\n",
       "      <td>C</td>\n",
       "      <td>False</td>\n",
       "      <td>B</td>\n",
       "      <td>False</td>\n",
       "      <td>B</td>\n",
       "      <td>False</td>\n",
       "      <td>C</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2</td>\n",
       "      <td>You are playing Russian roulette with a six-sh...</td>\n",
       "      <td>True</td>\n",
       "      <td>D</td>\n",
       "      <td>D</td>\n",
       "      <td>True</td>\n",
       "      <td>D</td>\n",
       "      <td>False</td>\n",
       "      <td>B</td>\n",
       "      <td>False</td>\n",
       "      <td>...</td>\n",
       "      <td>True</td>\n",
       "      <td>D</td>\n",
       "      <td>False</td>\n",
       "      <td>A</td>\n",
       "      <td>True</td>\n",
       "      <td>D</td>\n",
       "      <td>False</td>\n",
       "      <td>B</td>\n",
       "      <td>False</td>\n",
       "      <td>B</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>3</td>\n",
       "      <td>A farmer wants to cross a river and take with ...</td>\n",
       "      <td>True</td>\n",
       "      <td>B</td>\n",
       "      <td>B</td>\n",
       "      <td>False</td>\n",
       "      <td>C</td>\n",
       "      <td>False</td>\n",
       "      <td>A</td>\n",
       "      <td>False</td>\n",
       "      <td>...</td>\n",
       "      <td>False</td>\n",
       "      <td>A</td>\n",
       "      <td>False</td>\n",
       "      <td>A</td>\n",
       "      <td>False</td>\n",
       "      <td>C</td>\n",
       "      <td>False</td>\n",
       "      <td>C</td>\n",
       "      <td>False</td>\n",
       "      <td>A</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>4</td>\n",
       "      <td>Bob has three boxes in front of him - Box A, B...</td>\n",
       "      <td>True</td>\n",
       "      <td>A</td>\n",
       "      <td>A</td>\n",
       "      <td>False</td>\n",
       "      <td>C</td>\n",
       "      <td>True</td>\n",
       "      <td>A</td>\n",
       "      <td>True</td>\n",
       "      <td>...</td>\n",
       "      <td>True</td>\n",
       "      <td>A</td>\n",
       "      <td>False</td>\n",
       "      <td>C</td>\n",
       "      <td>False</td>\n",
       "      <td>C</td>\n",
       "      <td>False</td>\n",
       "      <td>C</td>\n",
       "      <td>True</td>\n",
       "      <td>A</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>295</th>\n",
       "      <td>295</td>\n",
       "      <td>A runaway trolley is heading down the tracks a...</td>\n",
       "      <td>True</td>\n",
       "      <td>A</td>\n",
       "      <td>A</td>\n",
       "      <td>False</td>\n",
       "      <td>B</td>\n",
       "      <td>False</td>\n",
       "      <td>C</td>\n",
       "      <td>False</td>\n",
       "      <td>...</td>\n",
       "      <td>True</td>\n",
       "      <td>A</td>\n",
       "      <td>False</td>\n",
       "      <td>B</td>\n",
       "      <td>False</td>\n",
       "      <td>B</td>\n",
       "      <td>False</td>\n",
       "      <td>B</td>\n",
       "      <td>False</td>\n",
       "      <td>B</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>296</th>\n",
       "      <td>296</td>\n",
       "      <td>How do you efficiently measure exactly 4 gallo...</td>\n",
       "      <td>False</td>\n",
       "      <td>A</td>\n",
       "      <td>B</td>\n",
       "      <td>True</td>\n",
       "      <td>A</td>\n",
       "      <td>True</td>\n",
       "      <td>A</td>\n",
       "      <td>False</td>\n",
       "      <td>...</td>\n",
       "      <td>False</td>\n",
       "      <td>D</td>\n",
       "      <td>False</td>\n",
       "      <td>B</td>\n",
       "      <td>False</td>\n",
       "      <td>D</td>\n",
       "      <td>False</td>\n",
       "      <td>D</td>\n",
       "      <td>True</td>\n",
       "      <td>A</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>297</th>\n",
       "      <td>297</td>\n",
       "      <td>A 2kg tree grows in a planted pot with 10kg of...</td>\n",
       "      <td>False</td>\n",
       "      <td>A</td>\n",
       "      <td>D</td>\n",
       "      <td>True</td>\n",
       "      <td>A</td>\n",
       "      <td>False</td>\n",
       "      <td>D</td>\n",
       "      <td>True</td>\n",
       "      <td>...</td>\n",
       "      <td>False</td>\n",
       "      <td>D</td>\n",
       "      <td>False</td>\n",
       "      <td>D</td>\n",
       "      <td>False</td>\n",
       "      <td>D</td>\n",
       "      <td>True</td>\n",
       "      <td>A</td>\n",
       "      <td>True</td>\n",
       "      <td>A</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>298</th>\n",
       "      <td>298</td>\n",
       "      <td>A man steps over a 0.5m hurdle. What is the mi...</td>\n",
       "      <td>False</td>\n",
       "      <td>B</td>\n",
       "      <td>D</td>\n",
       "      <td>False</td>\n",
       "      <td>D</td>\n",
       "      <td>False</td>\n",
       "      <td>D</td>\n",
       "      <td>False</td>\n",
       "      <td>...</td>\n",
       "      <td>False</td>\n",
       "      <td>A</td>\n",
       "      <td>False</td>\n",
       "      <td>D</td>\n",
       "      <td>False</td>\n",
       "      <td>D</td>\n",
       "      <td>True</td>\n",
       "      <td>B</td>\n",
       "      <td>False</td>\n",
       "      <td>A</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>299</th>\n",
       "      <td>299</td>\n",
       "      <td>I walk to my friend's house, averaging 3mph. H...</td>\n",
       "      <td>True</td>\n",
       "      <td>D</td>\n",
       "      <td>D</td>\n",
       "      <td>True</td>\n",
       "      <td>D</td>\n",
       "      <td>False</td>\n",
       "      <td>C</td>\n",
       "      <td>True</td>\n",
       "      <td>...</td>\n",
       "      <td>True</td>\n",
       "      <td>D</td>\n",
       "      <td>True</td>\n",
       "      <td>D</td>\n",
       "      <td>True</td>\n",
       "      <td>D</td>\n",
       "      <td>True</td>\n",
       "      <td>D</td>\n",
       "      <td>False</td>\n",
       "      <td>None</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>300 rows × 27 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "    bench                                                     \\\n",
       "    index                                           question   \n",
       "0       0  You have six horses and want to race them to s...   \n",
       "1       1  Suppose you're on a game show, and you're give...   \n",
       "2       2  You are playing Russian roulette with a six-sh...   \n",
       "3       3  A farmer wants to cross a river and take with ...   \n",
       "4       4  Bob has three boxes in front of him - Box A, B...   \n",
       "..    ...                                                ...   \n",
       "295   295  A runaway trolley is heading down the tracks a...   \n",
       "296   296  How do you efficiently measure exactly 4 gallo...   \n",
       "297   297  A 2kg tree grows in a planted pot with 10kg of...   \n",
       "298   298  A man steps over a 0.5m hurdle. What is the mi...   \n",
       "299   299  I walk to my friend's house, averaging 3mph. H...   \n",
       "\n",
       "    claude-3-5-sonnet-20240620                     bench  \\\n",
       "                       correct normalized_correct_letter   \n",
       "0                        False                         A   \n",
       "1                         True                         D   \n",
       "2                         True                         D   \n",
       "3                         True                         B   \n",
       "4                         True                         A   \n",
       "..                         ...                       ...   \n",
       "295                       True                         A   \n",
       "296                      False                         A   \n",
       "297                      False                         A   \n",
       "298                      False                         B   \n",
       "299                       True                         D   \n",
       "\n",
       "    claude-3-5-sonnet-20240620 claude-3-opus-20240229  \\\n",
       "      normalized_choice_letter                correct   \n",
       "0                            C                  False   \n",
       "1                            D                  False   \n",
       "2                            D                   True   \n",
       "3                            B                  False   \n",
       "4                            A                  False   \n",
       "..                         ...                    ...   \n",
       "295                          A                  False   \n",
       "296                          B                   True   \n",
       "297                          D                   True   \n",
       "298                          D                  False   \n",
       "299                          D                   True   \n",
       "\n",
       "                             command-r-plus                           \\\n",
       "    normalized_choice_letter        correct normalized_choice_letter   \n",
       "0                          B           True                        A   \n",
       "1                          C          False                        B   \n",
       "2                          D          False                        B   \n",
       "3                          C          False                        A   \n",
       "4                          C           True                        A   \n",
       "..                       ...            ...                      ...   \n",
       "295                        B          False                        C   \n",
       "296                        A           True                        A   \n",
       "297                        A          False                        D   \n",
       "298                        D          False                        D   \n",
       "299                        D          False                        C   \n",
       "\n",
       "    gemini-1_5-pro  ...  \\\n",
       "           correct  ...   \n",
       "0            False  ...   \n",
       "1            False  ...   \n",
       "2            False  ...   \n",
       "3            False  ...   \n",
       "4             True  ...   \n",
       "..             ...  ...   \n",
       "295          False  ...   \n",
       "296          False  ...   \n",
       "297           True  ...   \n",
       "298          False  ...   \n",
       "299           True  ...   \n",
       "\n",
       "    Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com  \\\n",
       "                                                        correct   \n",
       "0                                                False            \n",
       "1                                                False            \n",
       "2                                                 True            \n",
       "3                                                False            \n",
       "4                                                 True            \n",
       "..                                                 ...            \n",
       "295                                               True            \n",
       "296                                              False            \n",
       "297                                              False            \n",
       "298                                              False            \n",
       "299                                               True            \n",
       "\n",
       "                              \\\n",
       "    normalized_choice_letter   \n",
       "0                          B   \n",
       "1                          B   \n",
       "2                          D   \n",
       "3                          A   \n",
       "4                          A   \n",
       "..                       ...   \n",
       "295                        A   \n",
       "296                        D   \n",
       "297                        D   \n",
       "298                        A   \n",
       "299                        D   \n",
       "\n",
       "    Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com  \\\n",
       "                                                        correct   \n",
       "0                                                 True            \n",
       "1                                                False            \n",
       "2                                                False            \n",
       "3                                                False            \n",
       "4                                                False            \n",
       "..                                                 ...            \n",
       "295                                              False            \n",
       "296                                              False            \n",
       "297                                              False            \n",
       "298                                              False            \n",
       "299                                               True            \n",
       "\n",
       "                             meta_llama3-70b-instruct-v1_0  \\\n",
       "    normalized_choice_letter                       correct   \n",
       "0                          A                         False   \n",
       "1                          C                         False   \n",
       "2                          A                          True   \n",
       "3                          A                         False   \n",
       "4                          C                         False   \n",
       "..                       ...                           ...   \n",
       "295                        B                         False   \n",
       "296                        B                         False   \n",
       "297                        D                         False   \n",
       "298                        D                         False   \n",
       "299                        D                          True   \n",
       "\n",
       "                             mistral-large-latest                           \\\n",
       "    normalized_choice_letter              correct normalized_choice_letter   \n",
       "0                          B                False                        B   \n",
       "1                          B                False                        B   \n",
       "2                          D                False                        B   \n",
       "3                          C                False                        C   \n",
       "4                          C                False                        C   \n",
       "..                       ...                  ...                      ...   \n",
       "295                        B                False                        B   \n",
       "296                        D                False                        D   \n",
       "297                        D                 True                        A   \n",
       "298                        D                 True                        B   \n",
       "299                        D                 True                        D   \n",
       "\n",
       "    open-mixtral-8x22b                           \n",
       "               correct normalized_choice_letter  \n",
       "0                False                        B  \n",
       "1                False                        C  \n",
       "2                False                        B  \n",
       "3                False                        A  \n",
       "4                 True                        A  \n",
       "..                 ...                      ...  \n",
       "295              False                        B  \n",
       "296               True                        A  \n",
       "297               True                        A  \n",
       "298              False                        A  \n",
       "299              False                     None  \n",
       "\n",
       "[300 rows x 27 columns]"
      ]
     },
     "execution_count": 47,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "moe_llm_answers_comb_df = {}\n",
    "for llm_model in list(all_llm_answers.keys()):\n",
    "    data = pd.DataFrame(moe_llm_answers[llm_model])\n",
    "    data.reset_index(drop=False, inplace=True)\n",
    "    columns_mapping = {\n",
    "        'index': ('bench', 'index'),\n",
    "        'question': ('bench', 'question'),\n",
    "        'normalized_correct_letter': ('bench', 'normalized_correct_letter'),\n",
    "        'normalized_choice_letter': (llm_model, 'normalized_choice_letter'),\n",
    "        'correct': (llm_model, 'correct'),\n",
    "    }\n",
    "    grouped_columns_map = [columns_mapping[col] for col in data.columns if col in columns_mapping]\n",
    "    grouped_columns = pd.MultiIndex.from_tuples(grouped_columns_map)\n",
    "    data.columns = grouped_columns\n",
    "    data\n",
    "    if len(moe_llm_answers_comb_df) == 0:\n",
    "        moe_llm_answers_comb_df = data\n",
    "    else:\n",
    "        moe_llm_answers_comb_df = pd.merge(moe_llm_answers_comb_df, data, on=[\n",
    "            ('bench', 'index'), ('bench', 'question'), ('bench', 'normalized_correct_letter')\n",
    "        ])\n",
    "\n",
    "assert len(moe_llm_answers_comb_df) == len(all_llm_answers[list(all_llm_answers.keys())[0]]), 'Unable to reliably merge based on index, question, and normalized_correct_letter. This might suggest that the data is inconsistent or inaccurate'\n",
    "\n",
    "moe_llm_answers_comb_df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 50,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[('claude-3-5-sonnet-20240620', 'normalized_choice_letter'),\n",
       " ('claude-3-opus-20240229', 'normalized_choice_letter'),\n",
       " ('command-r-plus', 'normalized_choice_letter'),\n",
       " ('gemini-1_5-pro', 'normalized_choice_letter'),\n",
       " ('gpt-4-turbo-preview', 'normalized_choice_letter'),\n",
       " ('gpt-4o-mini-2024-07-18', 'normalized_choice_letter'),\n",
       " ('gpt-4o', 'normalized_choice_letter'),\n",
       " ('Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com',\n",
       "  'normalized_choice_letter'),\n",
       " ('Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com',\n",
       "  'normalized_choice_letter'),\n",
       " ('meta_llama3-70b-instruct-v1_0', 'normalized_choice_letter'),\n",
       " ('mistral-large-latest', 'normalized_choice_letter'),\n",
       " ('open-mixtral-8x22b', 'normalized_choice_letter')]"
      ]
     },
     "execution_count": 50,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "llm_models = list(all_llm_answers.keys())\n",
    "norm_choice_cols = list(zip(llm_models, ['normalized_choice_letter'] * len(llm_models)))\n",
    "norm_correct_cols = list(zip(llm_models, ['correct'] * len(llm_models)))\n",
    "norm_choice_cols"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "import seaborn as sns\n",
    "import matplotlib.pyplot as plt"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 66,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 600x400 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "question_correct_df = moe_llm_answers_comb_df[norm_correct_cols + [('bench', 'question')]].groupby([('bench', 'question')]).median().reset_index(drop=True)\n",
    "question_correct_df.index += 1\n",
    "\n",
    "plt.figure(figsize=(6, 4))\n",
    "cmap = sns.color_palette(['red', 'green'])\n",
    "sns.heatmap(question_correct_df.T, cmap=cmap, cbar=False, linewidths=1, linecolor='grey')\n",
    "#plt.xticks(rotation=0, ha='right')\n",
    "None"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 49,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 600x400 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "answer_correlations = {}\n",
    "for (model_base, col) in norm_choice_cols:\n",
    "    model_correlations = {}\n",
    "    for (model_test, col) in norm_choice_cols:\n",
    "        same_answer_percentage = ((moe_llm_answers_comb_df[model_base][col].fillna(0) \n",
    "                                   == moe_llm_answers_comb_df[model_test][col].fillna(0))\n",
    "                                  .value_counts(normalize=True)[True])\n",
    "        model_correlations[model_test] = same_answer_percentage\n",
    "    answer_correlations[model_base] = model_correlations\n",
    "\n",
    "correlation_matrix = pd.DataFrame(answer_correlations)\n",
    "\n",
    "plt.figure(figsize=(6, 4))\n",
    "ax = sns.heatmap(correlation_matrix, annot=True, cmap='coolwarm', vmin=0, vmax=1, \n",
    "                 cbar_kws={\"shrink\": 0.5}, fmt='.2f', annot_kws={'size': 7})\n",
    "plt.xticks(rotation=45, ha='right')\n",
    "plt.title('Correlation Matrix')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Run test to see how MOE performs with various permutations of models"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com    0.546667\n",
       "claude-3-5-sonnet-20240620                                     0.506667\n",
       "claude-3-opus-20240229                                         0.443333\n",
       "gpt-4o                                                         0.420000\n",
       "gpt-4-turbo-preview                                            0.366667\n",
       "Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com    0.343333\n",
       "gpt-4o-mini-2024-07-18                                         0.333333\n",
       "gemini-1_5-pro                                                 0.310000\n",
       "mistral-large-latest                                           0.296667\n",
       "meta_llama3-70b-instruct-v1_0                                  0.290000\n",
       "open-mixtral-8x22b                                             0.270000\n",
       "command-r-plus                                                 0.186667\n",
       "dtype: float64"
      ]
     },
     "execution_count": 27,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Sanity check to see if model scores match those calculated earlier (because we have done a lot of messing about). \n",
    "# It can used as a basis to see if MOE is improving results or not.\n",
    "result_series = pd.Series({col[0]: (moe_llm_answers_comb_df[col] == moe_llm_answers_comb_df['bench']['normalized_correct_letter']).value_counts(normalize=True)[True] \n",
    " for col in norm_choice_cols})\n",
    "\n",
    "result_series.sort_values(ascending=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "\n",
    "def get_most_common_or_random(row):\n",
    "    # Get the value counts for the row, if all NaNs, return a random choice\n",
    "    value_counts = row.value_counts()\n",
    "    if value_counts.empty:  # Check if value_counts is empty (all values are NaN)\n",
    "        return np.random.choice(['A', 'B', 'C', 'D'])\n",
    "    else:\n",
    "        # Find the maximum count\n",
    "        max_count = value_counts.max()\n",
    "        # Find the indices (letters) that have the maximum count\n",
    "        most_common_choices = value_counts[value_counts == max_count].index.tolist()\n",
    "        # If there's more than one most frequent letter, pick one randomly\n",
    "        return np.random.choice(most_common_choices)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [],
   "source": [
    "# subset_norm_choice_cols = [col for col in norm_choice_cols if col[0] in [\n",
    "#                             'claude-3-5-sonnet-20240620', \n",
    "#                             'claude-3-opus-20240229', \n",
    "#                             'gemini-1_5-pro', \n",
    "#                             'gpt-4o', \n",
    "#                             'open-mixtral-8x22b',\n",
    "#                            ]]\n",
    "\n",
    "subset_norm_choice_cols = norm_choice_cols"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "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>A</th>\n",
       "      <th>B</th>\n",
       "      <th>C</th>\n",
       "      <th>D</th>\n",
       "      <th>correct_letter</th>\n",
       "      <th>most_common_answer</th>\n",
       "      <th>largest_common_answer</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>4.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>A</td>\n",
       "      <td>B</td>\n",
       "      <td>5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>NaN</td>\n",
       "      <td>7.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>D</td>\n",
       "      <td>B</td>\n",
       "      <td>7</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>6.0</td>\n",
       "      <td>D</td>\n",
       "      <td>D</td>\n",
       "      <td>6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>5.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>6.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>B</td>\n",
       "      <td>C</td>\n",
       "      <td>6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>5.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>7.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>A</td>\n",
       "      <td>C</td>\n",
       "      <td>7</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>295</th>\n",
       "      <td>2.0</td>\n",
       "      <td>8.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>A</td>\n",
       "      <td>B</td>\n",
       "      <td>8</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>296</th>\n",
       "      <td>4.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>4.0</td>\n",
       "      <td>A</td>\n",
       "      <td>B</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>297</th>\n",
       "      <td>7.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>5.0</td>\n",
       "      <td>A</td>\n",
       "      <td>A</td>\n",
       "      <td>7</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>298</th>\n",
       "      <td>2.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>9.0</td>\n",
       "      <td>B</td>\n",
       "      <td>D</td>\n",
       "      <td>9</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>299</th>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1.0</td>\n",
       "      <td>10.0</td>\n",
       "      <td>D</td>\n",
       "      <td>D</td>\n",
       "      <td>10</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>300 rows × 7 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "       A    B    C     D correct_letter most_common_answer  \\\n",
       "0    4.0  5.0  2.0   1.0              A                  B   \n",
       "1    NaN  7.0  3.0   2.0              D                  B   \n",
       "2    1.0  5.0  NaN   6.0              D                  D   \n",
       "3    5.0  1.0  6.0   NaN              B                  C   \n",
       "4    5.0  NaN  7.0   NaN              A                  C   \n",
       "..   ...  ...  ...   ...            ...                ...   \n",
       "295  2.0  8.0  1.0   1.0              A                  B   \n",
       "296  4.0  4.0  NaN   4.0              A                  B   \n",
       "297  7.0  NaN  NaN   5.0              A                  A   \n",
       "298  2.0  1.0  NaN   9.0              B                  D   \n",
       "299  NaN  NaN  1.0  10.0              D                  D   \n",
       "\n",
       "     largest_common_answer  \n",
       "0                        5  \n",
       "1                        7  \n",
       "2                        6  \n",
       "3                        6  \n",
       "4                        7  \n",
       "..                     ...  \n",
       "295                      8  \n",
       "296                      4  \n",
       "297                      7  \n",
       "298                      9  \n",
       "299                     10  \n",
       "\n",
       "[300 rows x 7 columns]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Score %: 0.3933333333333333\n"
     ]
    }
   ],
   "source": [
    "\n",
    "answer_values = moe_llm_answers_comb_df.apply(axis='columns', \n",
    "    func=lambda x: x[subset_norm_choice_cols].value_counts())\n",
    "most_common_answer = moe_llm_answers_comb_df.apply(axis='columns', \n",
    "    func=lambda x: get_most_common_or_random(x[subset_norm_choice_cols]), result_type='expand')\n",
    "largest_common_answer = moe_llm_answers_comb_df.apply(axis='columns', \n",
    "    func=lambda x: x[subset_norm_choice_cols].value_counts().max())\n",
    "answer_values['correct_letter'] = moe_llm_answers_comb_df['bench']['normalized_correct_letter']\n",
    "answer_values['most_common_answer'] = most_common_answer\n",
    "answer_values['largest_common_answer'] = largest_common_answer\n",
    "display(answer_values)\n",
    "\n",
    "correct = (answer_values['correct_letter'] == answer_values['most_common_answer']).sum() / len(answer_values)\n",
    "print('Score %:', correct)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.3933333333333333\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<Axes: title={'center': 'Accuracy by Most Common Answer'}, xlabel='largest_common_answer'>"
      ]
     },
     "execution_count": 31,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "answer_values['correct'] = (answer_values['correct_letter'] == answer_values['most_common_answer']).astype(int)\n",
    "answer_values[['largest_common_answer', 'correct']].groupby('largest_common_answer').mean().plot(kind='bar', title='Accuracy by Most Common Answer')\n",
    "answer_values[['largest_common_answer', 'correct']].groupby('largest_common_answer').count().plot(kind='bar', title='Accuracy by Most Common Answer')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[('claude-3-5-sonnet-20240620', 'claude-3-opus-20240229'),\n",
       " ('claude-3-5-sonnet-20240620', 'command-r-plus'),\n",
       " ('claude-3-5-sonnet-20240620', 'gemini-1_5-pro'),\n",
       " ('claude-3-5-sonnet-20240620', 'gpt-4-turbo-preview'),\n",
       " ('claude-3-5-sonnet-20240620', 'gpt-4o-mini-2024-07-18'),\n",
       " ('claude-3-5-sonnet-20240620', 'gpt-4o'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com'),\n",
       " ('claude-3-5-sonnet-20240620', 'meta_llama3-70b-instruct-v1_0'),\n",
       " ('claude-3-5-sonnet-20240620', 'mistral-large-latest'),\n",
       " ('claude-3-5-sonnet-20240620', 'open-mixtral-8x22b'),\n",
       " ('claude-3-opus-20240229', 'command-r-plus'),\n",
       " ('claude-3-opus-20240229', 'gemini-1_5-pro'),\n",
       " ('claude-3-opus-20240229', 'gpt-4-turbo-preview'),\n",
       " ('claude-3-opus-20240229', 'gpt-4o-mini-2024-07-18'),\n",
       " ('claude-3-opus-20240229', 'gpt-4o'),\n",
       " ('claude-3-opus-20240229',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com'),\n",
       " ('claude-3-opus-20240229',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com'),\n",
       " ('claude-3-opus-20240229', 'meta_llama3-70b-instruct-v1_0'),\n",
       " ('claude-3-opus-20240229', 'mistral-large-latest'),\n",
       " ('claude-3-opus-20240229', 'open-mixtral-8x22b'),\n",
       " ('command-r-plus', 'gemini-1_5-pro'),\n",
       " ('command-r-plus', 'gpt-4-turbo-preview'),\n",
       " ('command-r-plus', 'gpt-4o-mini-2024-07-18'),\n",
       " ('command-r-plus', 'gpt-4o'),\n",
       " ('command-r-plus',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com'),\n",
       " ('command-r-plus',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com'),\n",
       " ('command-r-plus', 'meta_llama3-70b-instruct-v1_0'),\n",
       " ('command-r-plus', 'mistral-large-latest'),\n",
       " ('command-r-plus', 'open-mixtral-8x22b'),\n",
       " ('gemini-1_5-pro', 'gpt-4-turbo-preview'),\n",
       " ('gemini-1_5-pro', 'gpt-4o-mini-2024-07-18'),\n",
       " ('gemini-1_5-pro', 'gpt-4o'),\n",
       " ('gemini-1_5-pro',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com'),\n",
       " ('gemini-1_5-pro',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com'),\n",
       " ('gemini-1_5-pro', 'meta_llama3-70b-instruct-v1_0'),\n",
       " ('gemini-1_5-pro', 'mistral-large-latest'),\n",
       " ('gemini-1_5-pro', 'open-mixtral-8x22b'),\n",
       " ('gpt-4-turbo-preview', 'gpt-4o-mini-2024-07-18'),\n",
       " ('gpt-4-turbo-preview', 'gpt-4o'),\n",
       " ('gpt-4-turbo-preview',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com'),\n",
       " ('gpt-4-turbo-preview',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com'),\n",
       " ('gpt-4-turbo-preview', 'meta_llama3-70b-instruct-v1_0'),\n",
       " ('gpt-4-turbo-preview', 'mistral-large-latest'),\n",
       " ('gpt-4-turbo-preview', 'open-mixtral-8x22b'),\n",
       " ('gpt-4o-mini-2024-07-18', 'gpt-4o'),\n",
       " ('gpt-4o-mini-2024-07-18',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com'),\n",
       " ('gpt-4o-mini-2024-07-18',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com'),\n",
       " ('gpt-4o-mini-2024-07-18', 'meta_llama3-70b-instruct-v1_0'),\n",
       " ('gpt-4o-mini-2024-07-18', 'mistral-large-latest'),\n",
       " ('gpt-4o-mini-2024-07-18', 'open-mixtral-8x22b'),\n",
       " ('gpt-4o', 'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com'),\n",
       " ('gpt-4o', 'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com'),\n",
       " ('gpt-4o', 'meta_llama3-70b-instruct-v1_0'),\n",
       " ('gpt-4o', 'mistral-large-latest'),\n",
       " ('gpt-4o', 'open-mixtral-8x22b'),\n",
       " ('Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com'),\n",
       " ('Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com',\n",
       "  'meta_llama3-70b-instruct-v1_0'),\n",
       " ('Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com',\n",
       "  'mistral-large-latest'),\n",
       " ('Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com',\n",
       "  'meta_llama3-70b-instruct-v1_0'),\n",
       " ('Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com',\n",
       "  'mistral-large-latest'),\n",
       " ('Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('meta_llama3-70b-instruct-v1_0', 'mistral-large-latest'),\n",
       " ('meta_llama3-70b-instruct-v1_0', 'open-mixtral-8x22b'),\n",
       " ('mistral-large-latest', 'open-mixtral-8x22b'),\n",
       " ('claude-3-5-sonnet-20240620', 'claude-3-opus-20240229', 'command-r-plus'),\n",
       " ('claude-3-5-sonnet-20240620', 'claude-3-opus-20240229', 'gemini-1_5-pro'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'claude-3-opus-20240229',\n",
       "  'gpt-4-turbo-preview'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'claude-3-opus-20240229',\n",
       "  'gpt-4o-mini-2024-07-18'),\n",
       " ('claude-3-5-sonnet-20240620', 'claude-3-opus-20240229', 'gpt-4o'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'claude-3-opus-20240229',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'claude-3-opus-20240229',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'claude-3-opus-20240229',\n",
       "  'meta_llama3-70b-instruct-v1_0'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'claude-3-opus-20240229',\n",
       "  'mistral-large-latest'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'claude-3-opus-20240229',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('claude-3-5-sonnet-20240620', 'command-r-plus', 'gemini-1_5-pro'),\n",
       " ('claude-3-5-sonnet-20240620', 'command-r-plus', 'gpt-4-turbo-preview'),\n",
       " ('claude-3-5-sonnet-20240620', 'command-r-plus', 'gpt-4o-mini-2024-07-18'),\n",
       " ('claude-3-5-sonnet-20240620', 'command-r-plus', 'gpt-4o'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'command-r-plus',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'command-r-plus',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'command-r-plus',\n",
       "  'meta_llama3-70b-instruct-v1_0'),\n",
       " ('claude-3-5-sonnet-20240620', 'command-r-plus', 'mistral-large-latest'),\n",
       " ('claude-3-5-sonnet-20240620', 'command-r-plus', 'open-mixtral-8x22b'),\n",
       " ('claude-3-5-sonnet-20240620', 'gemini-1_5-pro', 'gpt-4-turbo-preview'),\n",
       " ('claude-3-5-sonnet-20240620', 'gemini-1_5-pro', 'gpt-4o-mini-2024-07-18'),\n",
       " ('claude-3-5-sonnet-20240620', 'gemini-1_5-pro', 'gpt-4o'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'gemini-1_5-pro',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'gemini-1_5-pro',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'gemini-1_5-pro',\n",
       "  'meta_llama3-70b-instruct-v1_0'),\n",
       " ('claude-3-5-sonnet-20240620', 'gemini-1_5-pro', 'mistral-large-latest'),\n",
       " ('claude-3-5-sonnet-20240620', 'gemini-1_5-pro', 'open-mixtral-8x22b'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'gpt-4-turbo-preview',\n",
       "  'gpt-4o-mini-2024-07-18'),\n",
       " ('claude-3-5-sonnet-20240620', 'gpt-4-turbo-preview', 'gpt-4o'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'gpt-4-turbo-preview',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'gpt-4-turbo-preview',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'gpt-4-turbo-preview',\n",
       "  'meta_llama3-70b-instruct-v1_0'),\n",
       " ('claude-3-5-sonnet-20240620', 'gpt-4-turbo-preview', 'mistral-large-latest'),\n",
       " ('claude-3-5-sonnet-20240620', 'gpt-4-turbo-preview', 'open-mixtral-8x22b'),\n",
       " ('claude-3-5-sonnet-20240620', 'gpt-4o-mini-2024-07-18', 'gpt-4o'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'gpt-4o-mini-2024-07-18',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'gpt-4o-mini-2024-07-18',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'gpt-4o-mini-2024-07-18',\n",
       "  'meta_llama3-70b-instruct-v1_0'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'gpt-4o-mini-2024-07-18',\n",
       "  'mistral-large-latest'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'gpt-4o-mini-2024-07-18',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'gpt-4o',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'gpt-4o',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com'),\n",
       " ('claude-3-5-sonnet-20240620', 'gpt-4o', 'meta_llama3-70b-instruct-v1_0'),\n",
       " ('claude-3-5-sonnet-20240620', 'gpt-4o', 'mistral-large-latest'),\n",
       " ('claude-3-5-sonnet-20240620', 'gpt-4o', 'open-mixtral-8x22b'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com',\n",
       "  'meta_llama3-70b-instruct-v1_0'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com',\n",
       "  'mistral-large-latest'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com',\n",
       "  'meta_llama3-70b-instruct-v1_0'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com',\n",
       "  'mistral-large-latest'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'meta_llama3-70b-instruct-v1_0',\n",
       "  'mistral-large-latest'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'meta_llama3-70b-instruct-v1_0',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('claude-3-5-sonnet-20240620', 'mistral-large-latest', 'open-mixtral-8x22b'),\n",
       " ('claude-3-opus-20240229', 'command-r-plus', 'gemini-1_5-pro'),\n",
       " ('claude-3-opus-20240229', 'command-r-plus', 'gpt-4-turbo-preview'),\n",
       " ('claude-3-opus-20240229', 'command-r-plus', 'gpt-4o-mini-2024-07-18'),\n",
       " ('claude-3-opus-20240229', 'command-r-plus', 'gpt-4o'),\n",
       " ('claude-3-opus-20240229',\n",
       "  'command-r-plus',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com'),\n",
       " ('claude-3-opus-20240229',\n",
       "  'command-r-plus',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com'),\n",
       " ('claude-3-opus-20240229', 'command-r-plus', 'meta_llama3-70b-instruct-v1_0'),\n",
       " ('claude-3-opus-20240229', 'command-r-plus', 'mistral-large-latest'),\n",
       " ('claude-3-opus-20240229', 'command-r-plus', 'open-mixtral-8x22b'),\n",
       " ('claude-3-opus-20240229', 'gemini-1_5-pro', 'gpt-4-turbo-preview'),\n",
       " ('claude-3-opus-20240229', 'gemini-1_5-pro', 'gpt-4o-mini-2024-07-18'),\n",
       " ('claude-3-opus-20240229', 'gemini-1_5-pro', 'gpt-4o'),\n",
       " ('claude-3-opus-20240229',\n",
       "  'gemini-1_5-pro',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com'),\n",
       " ('claude-3-opus-20240229',\n",
       "  'gemini-1_5-pro',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com'),\n",
       " ('claude-3-opus-20240229', 'gemini-1_5-pro', 'meta_llama3-70b-instruct-v1_0'),\n",
       " ('claude-3-opus-20240229', 'gemini-1_5-pro', 'mistral-large-latest'),\n",
       " ('claude-3-opus-20240229', 'gemini-1_5-pro', 'open-mixtral-8x22b'),\n",
       " ('claude-3-opus-20240229', 'gpt-4-turbo-preview', 'gpt-4o-mini-2024-07-18'),\n",
       " ('claude-3-opus-20240229', 'gpt-4-turbo-preview', 'gpt-4o'),\n",
       " ('claude-3-opus-20240229',\n",
       "  'gpt-4-turbo-preview',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com'),\n",
       " ('claude-3-opus-20240229',\n",
       "  'gpt-4-turbo-preview',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com'),\n",
       " ('claude-3-opus-20240229',\n",
       "  'gpt-4-turbo-preview',\n",
       "  'meta_llama3-70b-instruct-v1_0'),\n",
       " ('claude-3-opus-20240229', 'gpt-4-turbo-preview', 'mistral-large-latest'),\n",
       " ('claude-3-opus-20240229', 'gpt-4-turbo-preview', 'open-mixtral-8x22b'),\n",
       " ('claude-3-opus-20240229', 'gpt-4o-mini-2024-07-18', 'gpt-4o'),\n",
       " ('claude-3-opus-20240229',\n",
       "  'gpt-4o-mini-2024-07-18',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com'),\n",
       " ('claude-3-opus-20240229',\n",
       "  'gpt-4o-mini-2024-07-18',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com'),\n",
       " ('claude-3-opus-20240229',\n",
       "  'gpt-4o-mini-2024-07-18',\n",
       "  'meta_llama3-70b-instruct-v1_0'),\n",
       " ('claude-3-opus-20240229', 'gpt-4o-mini-2024-07-18', 'mistral-large-latest'),\n",
       " ('claude-3-opus-20240229', 'gpt-4o-mini-2024-07-18', 'open-mixtral-8x22b'),\n",
       " ('claude-3-opus-20240229',\n",
       "  'gpt-4o',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com'),\n",
       " ('claude-3-opus-20240229',\n",
       "  'gpt-4o',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com'),\n",
       " ('claude-3-opus-20240229', 'gpt-4o', 'meta_llama3-70b-instruct-v1_0'),\n",
       " ('claude-3-opus-20240229', 'gpt-4o', 'mistral-large-latest'),\n",
       " ('claude-3-opus-20240229', 'gpt-4o', 'open-mixtral-8x22b'),\n",
       " ('claude-3-opus-20240229',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com'),\n",
       " ('claude-3-opus-20240229',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com',\n",
       "  'meta_llama3-70b-instruct-v1_0'),\n",
       " ('claude-3-opus-20240229',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com',\n",
       "  'mistral-large-latest'),\n",
       " ('claude-3-opus-20240229',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('claude-3-opus-20240229',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com',\n",
       "  'meta_llama3-70b-instruct-v1_0'),\n",
       " ('claude-3-opus-20240229',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com',\n",
       "  'mistral-large-latest'),\n",
       " ('claude-3-opus-20240229',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('claude-3-opus-20240229',\n",
       "  'meta_llama3-70b-instruct-v1_0',\n",
       "  'mistral-large-latest'),\n",
       " ('claude-3-opus-20240229',\n",
       "  'meta_llama3-70b-instruct-v1_0',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('claude-3-opus-20240229', 'mistral-large-latest', 'open-mixtral-8x22b'),\n",
       " ('command-r-plus', 'gemini-1_5-pro', 'gpt-4-turbo-preview'),\n",
       " ('command-r-plus', 'gemini-1_5-pro', 'gpt-4o-mini-2024-07-18'),\n",
       " ('command-r-plus', 'gemini-1_5-pro', 'gpt-4o'),\n",
       " ('command-r-plus',\n",
       "  'gemini-1_5-pro',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com'),\n",
       " ('command-r-plus',\n",
       "  'gemini-1_5-pro',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com'),\n",
       " ('command-r-plus', 'gemini-1_5-pro', 'meta_llama3-70b-instruct-v1_0'),\n",
       " ('command-r-plus', 'gemini-1_5-pro', 'mistral-large-latest'),\n",
       " ('command-r-plus', 'gemini-1_5-pro', 'open-mixtral-8x22b'),\n",
       " ('command-r-plus', 'gpt-4-turbo-preview', 'gpt-4o-mini-2024-07-18'),\n",
       " ('command-r-plus', 'gpt-4-turbo-preview', 'gpt-4o'),\n",
       " ('command-r-plus',\n",
       "  'gpt-4-turbo-preview',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com'),\n",
       " ('command-r-plus',\n",
       "  'gpt-4-turbo-preview',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com'),\n",
       " ('command-r-plus', 'gpt-4-turbo-preview', 'meta_llama3-70b-instruct-v1_0'),\n",
       " ('command-r-plus', 'gpt-4-turbo-preview', 'mistral-large-latest'),\n",
       " ('command-r-plus', 'gpt-4-turbo-preview', 'open-mixtral-8x22b'),\n",
       " ('command-r-plus', 'gpt-4o-mini-2024-07-18', 'gpt-4o'),\n",
       " ('command-r-plus',\n",
       "  'gpt-4o-mini-2024-07-18',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com'),\n",
       " ('command-r-plus',\n",
       "  'gpt-4o-mini-2024-07-18',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com'),\n",
       " ('command-r-plus', 'gpt-4o-mini-2024-07-18', 'meta_llama3-70b-instruct-v1_0'),\n",
       " ('command-r-plus', 'gpt-4o-mini-2024-07-18', 'mistral-large-latest'),\n",
       " ('command-r-plus', 'gpt-4o-mini-2024-07-18', 'open-mixtral-8x22b'),\n",
       " ('command-r-plus',\n",
       "  'gpt-4o',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com'),\n",
       " ('command-r-plus',\n",
       "  'gpt-4o',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com'),\n",
       " ('command-r-plus', 'gpt-4o', 'meta_llama3-70b-instruct-v1_0'),\n",
       " ('command-r-plus', 'gpt-4o', 'mistral-large-latest'),\n",
       " ('command-r-plus', 'gpt-4o', 'open-mixtral-8x22b'),\n",
       " ('command-r-plus',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com'),\n",
       " ('command-r-plus',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com',\n",
       "  'meta_llama3-70b-instruct-v1_0'),\n",
       " ('command-r-plus',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com',\n",
       "  'mistral-large-latest'),\n",
       " ('command-r-plus',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('command-r-plus',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com',\n",
       "  'meta_llama3-70b-instruct-v1_0'),\n",
       " ('command-r-plus',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com',\n",
       "  'mistral-large-latest'),\n",
       " ('command-r-plus',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('command-r-plus', 'meta_llama3-70b-instruct-v1_0', 'mistral-large-latest'),\n",
       " ('command-r-plus', 'meta_llama3-70b-instruct-v1_0', 'open-mixtral-8x22b'),\n",
       " ('command-r-plus', 'mistral-large-latest', 'open-mixtral-8x22b'),\n",
       " ('gemini-1_5-pro', 'gpt-4-turbo-preview', 'gpt-4o-mini-2024-07-18'),\n",
       " ('gemini-1_5-pro', 'gpt-4-turbo-preview', 'gpt-4o'),\n",
       " ('gemini-1_5-pro',\n",
       "  'gpt-4-turbo-preview',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com'),\n",
       " ('gemini-1_5-pro',\n",
       "  'gpt-4-turbo-preview',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com'),\n",
       " ('gemini-1_5-pro', 'gpt-4-turbo-preview', 'meta_llama3-70b-instruct-v1_0'),\n",
       " ('gemini-1_5-pro', 'gpt-4-turbo-preview', 'mistral-large-latest'),\n",
       " ('gemini-1_5-pro', 'gpt-4-turbo-preview', 'open-mixtral-8x22b'),\n",
       " ('gemini-1_5-pro', 'gpt-4o-mini-2024-07-18', 'gpt-4o'),\n",
       " ('gemini-1_5-pro',\n",
       "  'gpt-4o-mini-2024-07-18',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com'),\n",
       " ('gemini-1_5-pro',\n",
       "  'gpt-4o-mini-2024-07-18',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com'),\n",
       " ('gemini-1_5-pro', 'gpt-4o-mini-2024-07-18', 'meta_llama3-70b-instruct-v1_0'),\n",
       " ('gemini-1_5-pro', 'gpt-4o-mini-2024-07-18', 'mistral-large-latest'),\n",
       " ('gemini-1_5-pro', 'gpt-4o-mini-2024-07-18', 'open-mixtral-8x22b'),\n",
       " ('gemini-1_5-pro',\n",
       "  'gpt-4o',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com'),\n",
       " ('gemini-1_5-pro',\n",
       "  'gpt-4o',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com'),\n",
       " ('gemini-1_5-pro', 'gpt-4o', 'meta_llama3-70b-instruct-v1_0'),\n",
       " ('gemini-1_5-pro', 'gpt-4o', 'mistral-large-latest'),\n",
       " ('gemini-1_5-pro', 'gpt-4o', 'open-mixtral-8x22b'),\n",
       " ('gemini-1_5-pro',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com'),\n",
       " ('gemini-1_5-pro',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com',\n",
       "  'meta_llama3-70b-instruct-v1_0'),\n",
       " ('gemini-1_5-pro',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com',\n",
       "  'mistral-large-latest'),\n",
       " ('gemini-1_5-pro',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('gemini-1_5-pro',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com',\n",
       "  'meta_llama3-70b-instruct-v1_0'),\n",
       " ('gemini-1_5-pro',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com',\n",
       "  'mistral-large-latest'),\n",
       " ('gemini-1_5-pro',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('gemini-1_5-pro', 'meta_llama3-70b-instruct-v1_0', 'mistral-large-latest'),\n",
       " ('gemini-1_5-pro', 'meta_llama3-70b-instruct-v1_0', 'open-mixtral-8x22b'),\n",
       " ('gemini-1_5-pro', 'mistral-large-latest', 'open-mixtral-8x22b'),\n",
       " ('gpt-4-turbo-preview', 'gpt-4o-mini-2024-07-18', 'gpt-4o'),\n",
       " ('gpt-4-turbo-preview',\n",
       "  'gpt-4o-mini-2024-07-18',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com'),\n",
       " ('gpt-4-turbo-preview',\n",
       "  'gpt-4o-mini-2024-07-18',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com'),\n",
       " ('gpt-4-turbo-preview',\n",
       "  'gpt-4o-mini-2024-07-18',\n",
       "  'meta_llama3-70b-instruct-v1_0'),\n",
       " ('gpt-4-turbo-preview', 'gpt-4o-mini-2024-07-18', 'mistral-large-latest'),\n",
       " ('gpt-4-turbo-preview', 'gpt-4o-mini-2024-07-18', 'open-mixtral-8x22b'),\n",
       " ('gpt-4-turbo-preview',\n",
       "  'gpt-4o',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com'),\n",
       " ('gpt-4-turbo-preview',\n",
       "  'gpt-4o',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com'),\n",
       " ('gpt-4-turbo-preview', 'gpt-4o', 'meta_llama3-70b-instruct-v1_0'),\n",
       " ('gpt-4-turbo-preview', 'gpt-4o', 'mistral-large-latest'),\n",
       " ('gpt-4-turbo-preview', 'gpt-4o', 'open-mixtral-8x22b'),\n",
       " ('gpt-4-turbo-preview',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com'),\n",
       " ('gpt-4-turbo-preview',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com',\n",
       "  'meta_llama3-70b-instruct-v1_0'),\n",
       " ('gpt-4-turbo-preview',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com',\n",
       "  'mistral-large-latest'),\n",
       " ('gpt-4-turbo-preview',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('gpt-4-turbo-preview',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com',\n",
       "  'meta_llama3-70b-instruct-v1_0'),\n",
       " ('gpt-4-turbo-preview',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com',\n",
       "  'mistral-large-latest'),\n",
       " ('gpt-4-turbo-preview',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('gpt-4-turbo-preview',\n",
       "  'meta_llama3-70b-instruct-v1_0',\n",
       "  'mistral-large-latest'),\n",
       " ('gpt-4-turbo-preview',\n",
       "  'meta_llama3-70b-instruct-v1_0',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('gpt-4-turbo-preview', 'mistral-large-latest', 'open-mixtral-8x22b'),\n",
       " ('gpt-4o-mini-2024-07-18',\n",
       "  'gpt-4o',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com'),\n",
       " ('gpt-4o-mini-2024-07-18',\n",
       "  'gpt-4o',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com'),\n",
       " ('gpt-4o-mini-2024-07-18', 'gpt-4o', 'meta_llama3-70b-instruct-v1_0'),\n",
       " ('gpt-4o-mini-2024-07-18', 'gpt-4o', 'mistral-large-latest'),\n",
       " ('gpt-4o-mini-2024-07-18', 'gpt-4o', 'open-mixtral-8x22b'),\n",
       " ('gpt-4o-mini-2024-07-18',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com'),\n",
       " ('gpt-4o-mini-2024-07-18',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com',\n",
       "  'meta_llama3-70b-instruct-v1_0'),\n",
       " ('gpt-4o-mini-2024-07-18',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com',\n",
       "  'mistral-large-latest'),\n",
       " ('gpt-4o-mini-2024-07-18',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('gpt-4o-mini-2024-07-18',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com',\n",
       "  'meta_llama3-70b-instruct-v1_0'),\n",
       " ('gpt-4o-mini-2024-07-18',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com',\n",
       "  'mistral-large-latest'),\n",
       " ('gpt-4o-mini-2024-07-18',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('gpt-4o-mini-2024-07-18',\n",
       "  'meta_llama3-70b-instruct-v1_0',\n",
       "  'mistral-large-latest'),\n",
       " ('gpt-4o-mini-2024-07-18',\n",
       "  'meta_llama3-70b-instruct-v1_0',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('gpt-4o-mini-2024-07-18', 'mistral-large-latest', 'open-mixtral-8x22b'),\n",
       " ('gpt-4o',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com'),\n",
       " ('gpt-4o',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com',\n",
       "  'meta_llama3-70b-instruct-v1_0'),\n",
       " ('gpt-4o',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com',\n",
       "  'mistral-large-latest'),\n",
       " ('gpt-4o',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('gpt-4o',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com',\n",
       "  'meta_llama3-70b-instruct-v1_0'),\n",
       " ('gpt-4o',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com',\n",
       "  'mistral-large-latest'),\n",
       " ('gpt-4o',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('gpt-4o', 'meta_llama3-70b-instruct-v1_0', 'mistral-large-latest'),\n",
       " ('gpt-4o', 'meta_llama3-70b-instruct-v1_0', 'open-mixtral-8x22b'),\n",
       " ('gpt-4o', 'mistral-large-latest', 'open-mixtral-8x22b'),\n",
       " ('Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com',\n",
       "  'meta_llama3-70b-instruct-v1_0'),\n",
       " ('Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com',\n",
       "  'mistral-large-latest'),\n",
       " ('Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com',\n",
       "  'meta_llama3-70b-instruct-v1_0',\n",
       "  'mistral-large-latest'),\n",
       " ('Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com',\n",
       "  'meta_llama3-70b-instruct-v1_0',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com',\n",
       "  'mistral-large-latest',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com',\n",
       "  'meta_llama3-70b-instruct-v1_0',\n",
       "  'mistral-large-latest'),\n",
       " ('Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com',\n",
       "  'meta_llama3-70b-instruct-v1_0',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com',\n",
       "  'mistral-large-latest',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('meta_llama3-70b-instruct-v1_0',\n",
       "  'mistral-large-latest',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'claude-3-opus-20240229',\n",
       "  'command-r-plus',\n",
       "  'gemini-1_5-pro'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'claude-3-opus-20240229',\n",
       "  'command-r-plus',\n",
       "  'gpt-4-turbo-preview'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'claude-3-opus-20240229',\n",
       "  'command-r-plus',\n",
       "  'gpt-4o-mini-2024-07-18'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'claude-3-opus-20240229',\n",
       "  'command-r-plus',\n",
       "  'gpt-4o'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'claude-3-opus-20240229',\n",
       "  'command-r-plus',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'claude-3-opus-20240229',\n",
       "  'command-r-plus',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'claude-3-opus-20240229',\n",
       "  'command-r-plus',\n",
       "  'meta_llama3-70b-instruct-v1_0'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'claude-3-opus-20240229',\n",
       "  'command-r-plus',\n",
       "  'mistral-large-latest'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'claude-3-opus-20240229',\n",
       "  'command-r-plus',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'claude-3-opus-20240229',\n",
       "  'gemini-1_5-pro',\n",
       "  'gpt-4-turbo-preview'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'claude-3-opus-20240229',\n",
       "  'gemini-1_5-pro',\n",
       "  'gpt-4o-mini-2024-07-18'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'claude-3-opus-20240229',\n",
       "  'gemini-1_5-pro',\n",
       "  'gpt-4o'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'claude-3-opus-20240229',\n",
       "  'gemini-1_5-pro',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'claude-3-opus-20240229',\n",
       "  'gemini-1_5-pro',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'claude-3-opus-20240229',\n",
       "  'gemini-1_5-pro',\n",
       "  'meta_llama3-70b-instruct-v1_0'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'claude-3-opus-20240229',\n",
       "  'gemini-1_5-pro',\n",
       "  'mistral-large-latest'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'claude-3-opus-20240229',\n",
       "  'gemini-1_5-pro',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'claude-3-opus-20240229',\n",
       "  'gpt-4-turbo-preview',\n",
       "  'gpt-4o-mini-2024-07-18'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'claude-3-opus-20240229',\n",
       "  'gpt-4-turbo-preview',\n",
       "  'gpt-4o'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'claude-3-opus-20240229',\n",
       "  'gpt-4-turbo-preview',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'claude-3-opus-20240229',\n",
       "  'gpt-4-turbo-preview',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'claude-3-opus-20240229',\n",
       "  'gpt-4-turbo-preview',\n",
       "  'meta_llama3-70b-instruct-v1_0'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'claude-3-opus-20240229',\n",
       "  'gpt-4-turbo-preview',\n",
       "  'mistral-large-latest'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'claude-3-opus-20240229',\n",
       "  'gpt-4-turbo-preview',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'claude-3-opus-20240229',\n",
       "  'gpt-4o-mini-2024-07-18',\n",
       "  'gpt-4o'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'claude-3-opus-20240229',\n",
       "  'gpt-4o-mini-2024-07-18',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'claude-3-opus-20240229',\n",
       "  'gpt-4o-mini-2024-07-18',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'claude-3-opus-20240229',\n",
       "  'gpt-4o-mini-2024-07-18',\n",
       "  'meta_llama3-70b-instruct-v1_0'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'claude-3-opus-20240229',\n",
       "  'gpt-4o-mini-2024-07-18',\n",
       "  'mistral-large-latest'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'claude-3-opus-20240229',\n",
       "  'gpt-4o-mini-2024-07-18',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'claude-3-opus-20240229',\n",
       "  'gpt-4o',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'claude-3-opus-20240229',\n",
       "  'gpt-4o',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'claude-3-opus-20240229',\n",
       "  'gpt-4o',\n",
       "  'meta_llama3-70b-instruct-v1_0'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'claude-3-opus-20240229',\n",
       "  'gpt-4o',\n",
       "  'mistral-large-latest'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'claude-3-opus-20240229',\n",
       "  'gpt-4o',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'claude-3-opus-20240229',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'claude-3-opus-20240229',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com',\n",
       "  'meta_llama3-70b-instruct-v1_0'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'claude-3-opus-20240229',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com',\n",
       "  'mistral-large-latest'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'claude-3-opus-20240229',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'claude-3-opus-20240229',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com',\n",
       "  'meta_llama3-70b-instruct-v1_0'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'claude-3-opus-20240229',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com',\n",
       "  'mistral-large-latest'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'claude-3-opus-20240229',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'claude-3-opus-20240229',\n",
       "  'meta_llama3-70b-instruct-v1_0',\n",
       "  'mistral-large-latest'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'claude-3-opus-20240229',\n",
       "  'meta_llama3-70b-instruct-v1_0',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'claude-3-opus-20240229',\n",
       "  'mistral-large-latest',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'command-r-plus',\n",
       "  'gemini-1_5-pro',\n",
       "  'gpt-4-turbo-preview'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'command-r-plus',\n",
       "  'gemini-1_5-pro',\n",
       "  'gpt-4o-mini-2024-07-18'),\n",
       " ('claude-3-5-sonnet-20240620', 'command-r-plus', 'gemini-1_5-pro', 'gpt-4o'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'command-r-plus',\n",
       "  'gemini-1_5-pro',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'command-r-plus',\n",
       "  'gemini-1_5-pro',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'command-r-plus',\n",
       "  'gemini-1_5-pro',\n",
       "  'meta_llama3-70b-instruct-v1_0'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'command-r-plus',\n",
       "  'gemini-1_5-pro',\n",
       "  'mistral-large-latest'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'command-r-plus',\n",
       "  'gemini-1_5-pro',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'command-r-plus',\n",
       "  'gpt-4-turbo-preview',\n",
       "  'gpt-4o-mini-2024-07-18'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'command-r-plus',\n",
       "  'gpt-4-turbo-preview',\n",
       "  'gpt-4o'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'command-r-plus',\n",
       "  'gpt-4-turbo-preview',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'command-r-plus',\n",
       "  'gpt-4-turbo-preview',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'command-r-plus',\n",
       "  'gpt-4-turbo-preview',\n",
       "  'meta_llama3-70b-instruct-v1_0'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'command-r-plus',\n",
       "  'gpt-4-turbo-preview',\n",
       "  'mistral-large-latest'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'command-r-plus',\n",
       "  'gpt-4-turbo-preview',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'command-r-plus',\n",
       "  'gpt-4o-mini-2024-07-18',\n",
       "  'gpt-4o'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'command-r-plus',\n",
       "  'gpt-4o-mini-2024-07-18',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'command-r-plus',\n",
       "  'gpt-4o-mini-2024-07-18',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'command-r-plus',\n",
       "  'gpt-4o-mini-2024-07-18',\n",
       "  'meta_llama3-70b-instruct-v1_0'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'command-r-plus',\n",
       "  'gpt-4o-mini-2024-07-18',\n",
       "  'mistral-large-latest'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'command-r-plus',\n",
       "  'gpt-4o-mini-2024-07-18',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'command-r-plus',\n",
       "  'gpt-4o',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'command-r-plus',\n",
       "  'gpt-4o',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'command-r-plus',\n",
       "  'gpt-4o',\n",
       "  'meta_llama3-70b-instruct-v1_0'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'command-r-plus',\n",
       "  'gpt-4o',\n",
       "  'mistral-large-latest'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'command-r-plus',\n",
       "  'gpt-4o',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'command-r-plus',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'command-r-plus',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com',\n",
       "  'meta_llama3-70b-instruct-v1_0'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'command-r-plus',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com',\n",
       "  'mistral-large-latest'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'command-r-plus',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'command-r-plus',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com',\n",
       "  'meta_llama3-70b-instruct-v1_0'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'command-r-plus',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com',\n",
       "  'mistral-large-latest'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'command-r-plus',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'command-r-plus',\n",
       "  'meta_llama3-70b-instruct-v1_0',\n",
       "  'mistral-large-latest'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'command-r-plus',\n",
       "  'meta_llama3-70b-instruct-v1_0',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'command-r-plus',\n",
       "  'mistral-large-latest',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'gemini-1_5-pro',\n",
       "  'gpt-4-turbo-preview',\n",
       "  'gpt-4o-mini-2024-07-18'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'gemini-1_5-pro',\n",
       "  'gpt-4-turbo-preview',\n",
       "  'gpt-4o'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'gemini-1_5-pro',\n",
       "  'gpt-4-turbo-preview',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'gemini-1_5-pro',\n",
       "  'gpt-4-turbo-preview',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'gemini-1_5-pro',\n",
       "  'gpt-4-turbo-preview',\n",
       "  'meta_llama3-70b-instruct-v1_0'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'gemini-1_5-pro',\n",
       "  'gpt-4-turbo-preview',\n",
       "  'mistral-large-latest'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'gemini-1_5-pro',\n",
       "  'gpt-4-turbo-preview',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'gemini-1_5-pro',\n",
       "  'gpt-4o-mini-2024-07-18',\n",
       "  'gpt-4o'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'gemini-1_5-pro',\n",
       "  'gpt-4o-mini-2024-07-18',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'gemini-1_5-pro',\n",
       "  'gpt-4o-mini-2024-07-18',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'gemini-1_5-pro',\n",
       "  'gpt-4o-mini-2024-07-18',\n",
       "  'meta_llama3-70b-instruct-v1_0'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'gemini-1_5-pro',\n",
       "  'gpt-4o-mini-2024-07-18',\n",
       "  'mistral-large-latest'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'gemini-1_5-pro',\n",
       "  'gpt-4o-mini-2024-07-18',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'gemini-1_5-pro',\n",
       "  'gpt-4o',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'gemini-1_5-pro',\n",
       "  'gpt-4o',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'gemini-1_5-pro',\n",
       "  'gpt-4o',\n",
       "  'meta_llama3-70b-instruct-v1_0'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'gemini-1_5-pro',\n",
       "  'gpt-4o',\n",
       "  'mistral-large-latest'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'gemini-1_5-pro',\n",
       "  'gpt-4o',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'gemini-1_5-pro',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'gemini-1_5-pro',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com',\n",
       "  'meta_llama3-70b-instruct-v1_0'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'gemini-1_5-pro',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com',\n",
       "  'mistral-large-latest'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'gemini-1_5-pro',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'gemini-1_5-pro',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com',\n",
       "  'meta_llama3-70b-instruct-v1_0'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'gemini-1_5-pro',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com',\n",
       "  'mistral-large-latest'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'gemini-1_5-pro',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'gemini-1_5-pro',\n",
       "  'meta_llama3-70b-instruct-v1_0',\n",
       "  'mistral-large-latest'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'gemini-1_5-pro',\n",
       "  'meta_llama3-70b-instruct-v1_0',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'gemini-1_5-pro',\n",
       "  'mistral-large-latest',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'gpt-4-turbo-preview',\n",
       "  'gpt-4o-mini-2024-07-18',\n",
       "  'gpt-4o'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'gpt-4-turbo-preview',\n",
       "  'gpt-4o-mini-2024-07-18',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'gpt-4-turbo-preview',\n",
       "  'gpt-4o-mini-2024-07-18',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'gpt-4-turbo-preview',\n",
       "  'gpt-4o-mini-2024-07-18',\n",
       "  'meta_llama3-70b-instruct-v1_0'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'gpt-4-turbo-preview',\n",
       "  'gpt-4o-mini-2024-07-18',\n",
       "  'mistral-large-latest'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'gpt-4-turbo-preview',\n",
       "  'gpt-4o-mini-2024-07-18',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'gpt-4-turbo-preview',\n",
       "  'gpt-4o',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'gpt-4-turbo-preview',\n",
       "  'gpt-4o',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'gpt-4-turbo-preview',\n",
       "  'gpt-4o',\n",
       "  'meta_llama3-70b-instruct-v1_0'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'gpt-4-turbo-preview',\n",
       "  'gpt-4o',\n",
       "  'mistral-large-latest'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'gpt-4-turbo-preview',\n",
       "  'gpt-4o',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'gpt-4-turbo-preview',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'gpt-4-turbo-preview',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com',\n",
       "  'meta_llama3-70b-instruct-v1_0'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'gpt-4-turbo-preview',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com',\n",
       "  'mistral-large-latest'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'gpt-4-turbo-preview',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'gpt-4-turbo-preview',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com',\n",
       "  'meta_llama3-70b-instruct-v1_0'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'gpt-4-turbo-preview',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com',\n",
       "  'mistral-large-latest'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'gpt-4-turbo-preview',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'gpt-4-turbo-preview',\n",
       "  'meta_llama3-70b-instruct-v1_0',\n",
       "  'mistral-large-latest'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'gpt-4-turbo-preview',\n",
       "  'meta_llama3-70b-instruct-v1_0',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'gpt-4-turbo-preview',\n",
       "  'mistral-large-latest',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'gpt-4o-mini-2024-07-18',\n",
       "  'gpt-4o',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'gpt-4o-mini-2024-07-18',\n",
       "  'gpt-4o',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'gpt-4o-mini-2024-07-18',\n",
       "  'gpt-4o',\n",
       "  'meta_llama3-70b-instruct-v1_0'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'gpt-4o-mini-2024-07-18',\n",
       "  'gpt-4o',\n",
       "  'mistral-large-latest'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'gpt-4o-mini-2024-07-18',\n",
       "  'gpt-4o',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'gpt-4o-mini-2024-07-18',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'gpt-4o-mini-2024-07-18',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com',\n",
       "  'meta_llama3-70b-instruct-v1_0'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'gpt-4o-mini-2024-07-18',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com',\n",
       "  'mistral-large-latest'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'gpt-4o-mini-2024-07-18',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'gpt-4o-mini-2024-07-18',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com',\n",
       "  'meta_llama3-70b-instruct-v1_0'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'gpt-4o-mini-2024-07-18',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com',\n",
       "  'mistral-large-latest'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'gpt-4o-mini-2024-07-18',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'gpt-4o-mini-2024-07-18',\n",
       "  'meta_llama3-70b-instruct-v1_0',\n",
       "  'mistral-large-latest'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'gpt-4o-mini-2024-07-18',\n",
       "  'meta_llama3-70b-instruct-v1_0',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'gpt-4o-mini-2024-07-18',\n",
       "  'mistral-large-latest',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'gpt-4o',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'gpt-4o',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com',\n",
       "  'meta_llama3-70b-instruct-v1_0'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'gpt-4o',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com',\n",
       "  'mistral-large-latest'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'gpt-4o',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'gpt-4o',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com',\n",
       "  'meta_llama3-70b-instruct-v1_0'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'gpt-4o',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com',\n",
       "  'mistral-large-latest'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'gpt-4o',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'gpt-4o',\n",
       "  'meta_llama3-70b-instruct-v1_0',\n",
       "  'mistral-large-latest'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'gpt-4o',\n",
       "  'meta_llama3-70b-instruct-v1_0',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'gpt-4o',\n",
       "  'mistral-large-latest',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com',\n",
       "  'meta_llama3-70b-instruct-v1_0'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com',\n",
       "  'mistral-large-latest'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com',\n",
       "  'meta_llama3-70b-instruct-v1_0',\n",
       "  'mistral-large-latest'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com',\n",
       "  'meta_llama3-70b-instruct-v1_0',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com',\n",
       "  'mistral-large-latest',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com',\n",
       "  'meta_llama3-70b-instruct-v1_0',\n",
       "  'mistral-large-latest'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com',\n",
       "  'meta_llama3-70b-instruct-v1_0',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com',\n",
       "  'mistral-large-latest',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'meta_llama3-70b-instruct-v1_0',\n",
       "  'mistral-large-latest',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('claude-3-opus-20240229',\n",
       "  'command-r-plus',\n",
       "  'gemini-1_5-pro',\n",
       "  'gpt-4-turbo-preview'),\n",
       " ('claude-3-opus-20240229',\n",
       "  'command-r-plus',\n",
       "  'gemini-1_5-pro',\n",
       "  'gpt-4o-mini-2024-07-18'),\n",
       " ('claude-3-opus-20240229', 'command-r-plus', 'gemini-1_5-pro', 'gpt-4o'),\n",
       " ('claude-3-opus-20240229',\n",
       "  'command-r-plus',\n",
       "  'gemini-1_5-pro',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com'),\n",
       " ('claude-3-opus-20240229',\n",
       "  'command-r-plus',\n",
       "  'gemini-1_5-pro',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com'),\n",
       " ('claude-3-opus-20240229',\n",
       "  'command-r-plus',\n",
       "  'gemini-1_5-pro',\n",
       "  'meta_llama3-70b-instruct-v1_0'),\n",
       " ('claude-3-opus-20240229',\n",
       "  'command-r-plus',\n",
       "  'gemini-1_5-pro',\n",
       "  'mistral-large-latest'),\n",
       " ('claude-3-opus-20240229',\n",
       "  'command-r-plus',\n",
       "  'gemini-1_5-pro',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('claude-3-opus-20240229',\n",
       "  'command-r-plus',\n",
       "  'gpt-4-turbo-preview',\n",
       "  'gpt-4o-mini-2024-07-18'),\n",
       " ('claude-3-opus-20240229', 'command-r-plus', 'gpt-4-turbo-preview', 'gpt-4o'),\n",
       " ('claude-3-opus-20240229',\n",
       "  'command-r-plus',\n",
       "  'gpt-4-turbo-preview',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com'),\n",
       " ('claude-3-opus-20240229',\n",
       "  'command-r-plus',\n",
       "  'gpt-4-turbo-preview',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com'),\n",
       " ('claude-3-opus-20240229',\n",
       "  'command-r-plus',\n",
       "  'gpt-4-turbo-preview',\n",
       "  'meta_llama3-70b-instruct-v1_0'),\n",
       " ('claude-3-opus-20240229',\n",
       "  'command-r-plus',\n",
       "  'gpt-4-turbo-preview',\n",
       "  'mistral-large-latest'),\n",
       " ('claude-3-opus-20240229',\n",
       "  'command-r-plus',\n",
       "  'gpt-4-turbo-preview',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('claude-3-opus-20240229',\n",
       "  'command-r-plus',\n",
       "  'gpt-4o-mini-2024-07-18',\n",
       "  'gpt-4o'),\n",
       " ('claude-3-opus-20240229',\n",
       "  'command-r-plus',\n",
       "  'gpt-4o-mini-2024-07-18',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com'),\n",
       " ('claude-3-opus-20240229',\n",
       "  'command-r-plus',\n",
       "  'gpt-4o-mini-2024-07-18',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com'),\n",
       " ('claude-3-opus-20240229',\n",
       "  'command-r-plus',\n",
       "  'gpt-4o-mini-2024-07-18',\n",
       "  'meta_llama3-70b-instruct-v1_0'),\n",
       " ('claude-3-opus-20240229',\n",
       "  'command-r-plus',\n",
       "  'gpt-4o-mini-2024-07-18',\n",
       "  'mistral-large-latest'),\n",
       " ('claude-3-opus-20240229',\n",
       "  'command-r-plus',\n",
       "  'gpt-4o-mini-2024-07-18',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('claude-3-opus-20240229',\n",
       "  'command-r-plus',\n",
       "  'gpt-4o',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com'),\n",
       " ('claude-3-opus-20240229',\n",
       "  'command-r-plus',\n",
       "  'gpt-4o',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com'),\n",
       " ('claude-3-opus-20240229',\n",
       "  'command-r-plus',\n",
       "  'gpt-4o',\n",
       "  'meta_llama3-70b-instruct-v1_0'),\n",
       " ('claude-3-opus-20240229',\n",
       "  'command-r-plus',\n",
       "  'gpt-4o',\n",
       "  'mistral-large-latest'),\n",
       " ('claude-3-opus-20240229', 'command-r-plus', 'gpt-4o', 'open-mixtral-8x22b'),\n",
       " ('claude-3-opus-20240229',\n",
       "  'command-r-plus',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com'),\n",
       " ('claude-3-opus-20240229',\n",
       "  'command-r-plus',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com',\n",
       "  'meta_llama3-70b-instruct-v1_0'),\n",
       " ('claude-3-opus-20240229',\n",
       "  'command-r-plus',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com',\n",
       "  'mistral-large-latest'),\n",
       " ('claude-3-opus-20240229',\n",
       "  'command-r-plus',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('claude-3-opus-20240229',\n",
       "  'command-r-plus',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com',\n",
       "  'meta_llama3-70b-instruct-v1_0'),\n",
       " ('claude-3-opus-20240229',\n",
       "  'command-r-plus',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com',\n",
       "  'mistral-large-latest'),\n",
       " ('claude-3-opus-20240229',\n",
       "  'command-r-plus',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('claude-3-opus-20240229',\n",
       "  'command-r-plus',\n",
       "  'meta_llama3-70b-instruct-v1_0',\n",
       "  'mistral-large-latest'),\n",
       " ('claude-3-opus-20240229',\n",
       "  'command-r-plus',\n",
       "  'meta_llama3-70b-instruct-v1_0',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('claude-3-opus-20240229',\n",
       "  'command-r-plus',\n",
       "  'mistral-large-latest',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('claude-3-opus-20240229',\n",
       "  'gemini-1_5-pro',\n",
       "  'gpt-4-turbo-preview',\n",
       "  'gpt-4o-mini-2024-07-18'),\n",
       " ('claude-3-opus-20240229', 'gemini-1_5-pro', 'gpt-4-turbo-preview', 'gpt-4o'),\n",
       " ('claude-3-opus-20240229',\n",
       "  'gemini-1_5-pro',\n",
       "  'gpt-4-turbo-preview',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com'),\n",
       " ('claude-3-opus-20240229',\n",
       "  'gemini-1_5-pro',\n",
       "  'gpt-4-turbo-preview',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com'),\n",
       " ('claude-3-opus-20240229',\n",
       "  'gemini-1_5-pro',\n",
       "  'gpt-4-turbo-preview',\n",
       "  'meta_llama3-70b-instruct-v1_0'),\n",
       " ('claude-3-opus-20240229',\n",
       "  'gemini-1_5-pro',\n",
       "  'gpt-4-turbo-preview',\n",
       "  'mistral-large-latest'),\n",
       " ('claude-3-opus-20240229',\n",
       "  'gemini-1_5-pro',\n",
       "  'gpt-4-turbo-preview',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('claude-3-opus-20240229',\n",
       "  'gemini-1_5-pro',\n",
       "  'gpt-4o-mini-2024-07-18',\n",
       "  'gpt-4o'),\n",
       " ('claude-3-opus-20240229',\n",
       "  'gemini-1_5-pro',\n",
       "  'gpt-4o-mini-2024-07-18',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com'),\n",
       " ('claude-3-opus-20240229',\n",
       "  'gemini-1_5-pro',\n",
       "  'gpt-4o-mini-2024-07-18',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com'),\n",
       " ('claude-3-opus-20240229',\n",
       "  'gemini-1_5-pro',\n",
       "  'gpt-4o-mini-2024-07-18',\n",
       "  'meta_llama3-70b-instruct-v1_0'),\n",
       " ('claude-3-opus-20240229',\n",
       "  'gemini-1_5-pro',\n",
       "  'gpt-4o-mini-2024-07-18',\n",
       "  'mistral-large-latest'),\n",
       " ('claude-3-opus-20240229',\n",
       "  'gemini-1_5-pro',\n",
       "  'gpt-4o-mini-2024-07-18',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('claude-3-opus-20240229',\n",
       "  'gemini-1_5-pro',\n",
       "  'gpt-4o',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com'),\n",
       " ('claude-3-opus-20240229',\n",
       "  'gemini-1_5-pro',\n",
       "  'gpt-4o',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com'),\n",
       " ('claude-3-opus-20240229',\n",
       "  'gemini-1_5-pro',\n",
       "  'gpt-4o',\n",
       "  'meta_llama3-70b-instruct-v1_0'),\n",
       " ('claude-3-opus-20240229',\n",
       "  'gemini-1_5-pro',\n",
       "  'gpt-4o',\n",
       "  'mistral-large-latest'),\n",
       " ('claude-3-opus-20240229', 'gemini-1_5-pro', 'gpt-4o', 'open-mixtral-8x22b'),\n",
       " ('claude-3-opus-20240229',\n",
       "  'gemini-1_5-pro',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com'),\n",
       " ('claude-3-opus-20240229',\n",
       "  'gemini-1_5-pro',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com',\n",
       "  'meta_llama3-70b-instruct-v1_0'),\n",
       " ('claude-3-opus-20240229',\n",
       "  'gemini-1_5-pro',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com',\n",
       "  'mistral-large-latest'),\n",
       " ('claude-3-opus-20240229',\n",
       "  'gemini-1_5-pro',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('claude-3-opus-20240229',\n",
       "  'gemini-1_5-pro',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com',\n",
       "  'meta_llama3-70b-instruct-v1_0'),\n",
       " ('claude-3-opus-20240229',\n",
       "  'gemini-1_5-pro',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com',\n",
       "  'mistral-large-latest'),\n",
       " ('claude-3-opus-20240229',\n",
       "  'gemini-1_5-pro',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('claude-3-opus-20240229',\n",
       "  'gemini-1_5-pro',\n",
       "  'meta_llama3-70b-instruct-v1_0',\n",
       "  'mistral-large-latest'),\n",
       " ('claude-3-opus-20240229',\n",
       "  'gemini-1_5-pro',\n",
       "  'meta_llama3-70b-instruct-v1_0',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('claude-3-opus-20240229',\n",
       "  'gemini-1_5-pro',\n",
       "  'mistral-large-latest',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('claude-3-opus-20240229',\n",
       "  'gpt-4-turbo-preview',\n",
       "  'gpt-4o-mini-2024-07-18',\n",
       "  'gpt-4o'),\n",
       " ('claude-3-opus-20240229',\n",
       "  'gpt-4-turbo-preview',\n",
       "  'gpt-4o-mini-2024-07-18',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com'),\n",
       " ('claude-3-opus-20240229',\n",
       "  'gpt-4-turbo-preview',\n",
       "  'gpt-4o-mini-2024-07-18',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com'),\n",
       " ('claude-3-opus-20240229',\n",
       "  'gpt-4-turbo-preview',\n",
       "  'gpt-4o-mini-2024-07-18',\n",
       "  'meta_llama3-70b-instruct-v1_0'),\n",
       " ('claude-3-opus-20240229',\n",
       "  'gpt-4-turbo-preview',\n",
       "  'gpt-4o-mini-2024-07-18',\n",
       "  'mistral-large-latest'),\n",
       " ('claude-3-opus-20240229',\n",
       "  'gpt-4-turbo-preview',\n",
       "  'gpt-4o-mini-2024-07-18',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('claude-3-opus-20240229',\n",
       "  'gpt-4-turbo-preview',\n",
       "  'gpt-4o',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com'),\n",
       " ('claude-3-opus-20240229',\n",
       "  'gpt-4-turbo-preview',\n",
       "  'gpt-4o',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com'),\n",
       " ('claude-3-opus-20240229',\n",
       "  'gpt-4-turbo-preview',\n",
       "  'gpt-4o',\n",
       "  'meta_llama3-70b-instruct-v1_0'),\n",
       " ('claude-3-opus-20240229',\n",
       "  'gpt-4-turbo-preview',\n",
       "  'gpt-4o',\n",
       "  'mistral-large-latest'),\n",
       " ('claude-3-opus-20240229',\n",
       "  'gpt-4-turbo-preview',\n",
       "  'gpt-4o',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('claude-3-opus-20240229',\n",
       "  'gpt-4-turbo-preview',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com'),\n",
       " ('claude-3-opus-20240229',\n",
       "  'gpt-4-turbo-preview',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com',\n",
       "  'meta_llama3-70b-instruct-v1_0'),\n",
       " ('claude-3-opus-20240229',\n",
       "  'gpt-4-turbo-preview',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com',\n",
       "  'mistral-large-latest'),\n",
       " ('claude-3-opus-20240229',\n",
       "  'gpt-4-turbo-preview',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('claude-3-opus-20240229',\n",
       "  'gpt-4-turbo-preview',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com',\n",
       "  'meta_llama3-70b-instruct-v1_0'),\n",
       " ('claude-3-opus-20240229',\n",
       "  'gpt-4-turbo-preview',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com',\n",
       "  'mistral-large-latest'),\n",
       " ('claude-3-opus-20240229',\n",
       "  'gpt-4-turbo-preview',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('claude-3-opus-20240229',\n",
       "  'gpt-4-turbo-preview',\n",
       "  'meta_llama3-70b-instruct-v1_0',\n",
       "  'mistral-large-latest'),\n",
       " ('claude-3-opus-20240229',\n",
       "  'gpt-4-turbo-preview',\n",
       "  'meta_llama3-70b-instruct-v1_0',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('claude-3-opus-20240229',\n",
       "  'gpt-4-turbo-preview',\n",
       "  'mistral-large-latest',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('claude-3-opus-20240229',\n",
       "  'gpt-4o-mini-2024-07-18',\n",
       "  'gpt-4o',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com'),\n",
       " ('claude-3-opus-20240229',\n",
       "  'gpt-4o-mini-2024-07-18',\n",
       "  'gpt-4o',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com'),\n",
       " ('claude-3-opus-20240229',\n",
       "  'gpt-4o-mini-2024-07-18',\n",
       "  'gpt-4o',\n",
       "  'meta_llama3-70b-instruct-v1_0'),\n",
       " ('claude-3-opus-20240229',\n",
       "  'gpt-4o-mini-2024-07-18',\n",
       "  'gpt-4o',\n",
       "  'mistral-large-latest'),\n",
       " ('claude-3-opus-20240229',\n",
       "  'gpt-4o-mini-2024-07-18',\n",
       "  'gpt-4o',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('claude-3-opus-20240229',\n",
       "  'gpt-4o-mini-2024-07-18',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com'),\n",
       " ('claude-3-opus-20240229',\n",
       "  'gpt-4o-mini-2024-07-18',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com',\n",
       "  'meta_llama3-70b-instruct-v1_0'),\n",
       " ('claude-3-opus-20240229',\n",
       "  'gpt-4o-mini-2024-07-18',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com',\n",
       "  'mistral-large-latest'),\n",
       " ('claude-3-opus-20240229',\n",
       "  'gpt-4o-mini-2024-07-18',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('claude-3-opus-20240229',\n",
       "  'gpt-4o-mini-2024-07-18',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com',\n",
       "  'meta_llama3-70b-instruct-v1_0'),\n",
       " ('claude-3-opus-20240229',\n",
       "  'gpt-4o-mini-2024-07-18',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com',\n",
       "  'mistral-large-latest'),\n",
       " ('claude-3-opus-20240229',\n",
       "  'gpt-4o-mini-2024-07-18',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('claude-3-opus-20240229',\n",
       "  'gpt-4o-mini-2024-07-18',\n",
       "  'meta_llama3-70b-instruct-v1_0',\n",
       "  'mistral-large-latest'),\n",
       " ('claude-3-opus-20240229',\n",
       "  'gpt-4o-mini-2024-07-18',\n",
       "  'meta_llama3-70b-instruct-v1_0',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('claude-3-opus-20240229',\n",
       "  'gpt-4o-mini-2024-07-18',\n",
       "  'mistral-large-latest',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('claude-3-opus-20240229',\n",
       "  'gpt-4o',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com'),\n",
       " ('claude-3-opus-20240229',\n",
       "  'gpt-4o',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com',\n",
       "  'meta_llama3-70b-instruct-v1_0'),\n",
       " ('claude-3-opus-20240229',\n",
       "  'gpt-4o',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com',\n",
       "  'mistral-large-latest'),\n",
       " ('claude-3-opus-20240229',\n",
       "  'gpt-4o',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('claude-3-opus-20240229',\n",
       "  'gpt-4o',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com',\n",
       "  'meta_llama3-70b-instruct-v1_0'),\n",
       " ('claude-3-opus-20240229',\n",
       "  'gpt-4o',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com',\n",
       "  'mistral-large-latest'),\n",
       " ('claude-3-opus-20240229',\n",
       "  'gpt-4o',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('claude-3-opus-20240229',\n",
       "  'gpt-4o',\n",
       "  'meta_llama3-70b-instruct-v1_0',\n",
       "  'mistral-large-latest'),\n",
       " ('claude-3-opus-20240229',\n",
       "  'gpt-4o',\n",
       "  'meta_llama3-70b-instruct-v1_0',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('claude-3-opus-20240229',\n",
       "  'gpt-4o',\n",
       "  'mistral-large-latest',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('claude-3-opus-20240229',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com',\n",
       "  'meta_llama3-70b-instruct-v1_0'),\n",
       " ('claude-3-opus-20240229',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com',\n",
       "  'mistral-large-latest'),\n",
       " ('claude-3-opus-20240229',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('claude-3-opus-20240229',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com',\n",
       "  'meta_llama3-70b-instruct-v1_0',\n",
       "  'mistral-large-latest'),\n",
       " ('claude-3-opus-20240229',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com',\n",
       "  'meta_llama3-70b-instruct-v1_0',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('claude-3-opus-20240229',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com',\n",
       "  'mistral-large-latest',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('claude-3-opus-20240229',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com',\n",
       "  'meta_llama3-70b-instruct-v1_0',\n",
       "  'mistral-large-latest'),\n",
       " ('claude-3-opus-20240229',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com',\n",
       "  'meta_llama3-70b-instruct-v1_0',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('claude-3-opus-20240229',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com',\n",
       "  'mistral-large-latest',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('claude-3-opus-20240229',\n",
       "  'meta_llama3-70b-instruct-v1_0',\n",
       "  'mistral-large-latest',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('command-r-plus',\n",
       "  'gemini-1_5-pro',\n",
       "  'gpt-4-turbo-preview',\n",
       "  'gpt-4o-mini-2024-07-18'),\n",
       " ('command-r-plus', 'gemini-1_5-pro', 'gpt-4-turbo-preview', 'gpt-4o'),\n",
       " ('command-r-plus',\n",
       "  'gemini-1_5-pro',\n",
       "  'gpt-4-turbo-preview',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com'),\n",
       " ('command-r-plus',\n",
       "  'gemini-1_5-pro',\n",
       "  'gpt-4-turbo-preview',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com'),\n",
       " ('command-r-plus',\n",
       "  'gemini-1_5-pro',\n",
       "  'gpt-4-turbo-preview',\n",
       "  'meta_llama3-70b-instruct-v1_0'),\n",
       " ('command-r-plus',\n",
       "  'gemini-1_5-pro',\n",
       "  'gpt-4-turbo-preview',\n",
       "  'mistral-large-latest'),\n",
       " ('command-r-plus',\n",
       "  'gemini-1_5-pro',\n",
       "  'gpt-4-turbo-preview',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('command-r-plus', 'gemini-1_5-pro', 'gpt-4o-mini-2024-07-18', 'gpt-4o'),\n",
       " ('command-r-plus',\n",
       "  'gemini-1_5-pro',\n",
       "  'gpt-4o-mini-2024-07-18',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com'),\n",
       " ('command-r-plus',\n",
       "  'gemini-1_5-pro',\n",
       "  'gpt-4o-mini-2024-07-18',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com'),\n",
       " ('command-r-plus',\n",
       "  'gemini-1_5-pro',\n",
       "  'gpt-4o-mini-2024-07-18',\n",
       "  'meta_llama3-70b-instruct-v1_0'),\n",
       " ('command-r-plus',\n",
       "  'gemini-1_5-pro',\n",
       "  'gpt-4o-mini-2024-07-18',\n",
       "  'mistral-large-latest'),\n",
       " ('command-r-plus',\n",
       "  'gemini-1_5-pro',\n",
       "  'gpt-4o-mini-2024-07-18',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('command-r-plus',\n",
       "  'gemini-1_5-pro',\n",
       "  'gpt-4o',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com'),\n",
       " ('command-r-plus',\n",
       "  'gemini-1_5-pro',\n",
       "  'gpt-4o',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com'),\n",
       " ('command-r-plus',\n",
       "  'gemini-1_5-pro',\n",
       "  'gpt-4o',\n",
       "  'meta_llama3-70b-instruct-v1_0'),\n",
       " ('command-r-plus', 'gemini-1_5-pro', 'gpt-4o', 'mistral-large-latest'),\n",
       " ('command-r-plus', 'gemini-1_5-pro', 'gpt-4o', 'open-mixtral-8x22b'),\n",
       " ('command-r-plus',\n",
       "  'gemini-1_5-pro',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com'),\n",
       " ('command-r-plus',\n",
       "  'gemini-1_5-pro',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com',\n",
       "  'meta_llama3-70b-instruct-v1_0'),\n",
       " ('command-r-plus',\n",
       "  'gemini-1_5-pro',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com',\n",
       "  'mistral-large-latest'),\n",
       " ('command-r-plus',\n",
       "  'gemini-1_5-pro',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('command-r-plus',\n",
       "  'gemini-1_5-pro',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com',\n",
       "  'meta_llama3-70b-instruct-v1_0'),\n",
       " ('command-r-plus',\n",
       "  'gemini-1_5-pro',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com',\n",
       "  'mistral-large-latest'),\n",
       " ('command-r-plus',\n",
       "  'gemini-1_5-pro',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('command-r-plus',\n",
       "  'gemini-1_5-pro',\n",
       "  'meta_llama3-70b-instruct-v1_0',\n",
       "  'mistral-large-latest'),\n",
       " ('command-r-plus',\n",
       "  'gemini-1_5-pro',\n",
       "  'meta_llama3-70b-instruct-v1_0',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('command-r-plus',\n",
       "  'gemini-1_5-pro',\n",
       "  'mistral-large-latest',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('command-r-plus', 'gpt-4-turbo-preview', 'gpt-4o-mini-2024-07-18', 'gpt-4o'),\n",
       " ('command-r-plus',\n",
       "  'gpt-4-turbo-preview',\n",
       "  'gpt-4o-mini-2024-07-18',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com'),\n",
       " ('command-r-plus',\n",
       "  'gpt-4-turbo-preview',\n",
       "  'gpt-4o-mini-2024-07-18',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com'),\n",
       " ('command-r-plus',\n",
       "  'gpt-4-turbo-preview',\n",
       "  'gpt-4o-mini-2024-07-18',\n",
       "  'meta_llama3-70b-instruct-v1_0'),\n",
       " ('command-r-plus',\n",
       "  'gpt-4-turbo-preview',\n",
       "  'gpt-4o-mini-2024-07-18',\n",
       "  'mistral-large-latest'),\n",
       " ('command-r-plus',\n",
       "  'gpt-4-turbo-preview',\n",
       "  'gpt-4o-mini-2024-07-18',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('command-r-plus',\n",
       "  'gpt-4-turbo-preview',\n",
       "  'gpt-4o',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com'),\n",
       " ('command-r-plus',\n",
       "  'gpt-4-turbo-preview',\n",
       "  'gpt-4o',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com'),\n",
       " ('command-r-plus',\n",
       "  'gpt-4-turbo-preview',\n",
       "  'gpt-4o',\n",
       "  'meta_llama3-70b-instruct-v1_0'),\n",
       " ('command-r-plus', 'gpt-4-turbo-preview', 'gpt-4o', 'mistral-large-latest'),\n",
       " ('command-r-plus', 'gpt-4-turbo-preview', 'gpt-4o', 'open-mixtral-8x22b'),\n",
       " ('command-r-plus',\n",
       "  'gpt-4-turbo-preview',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com'),\n",
       " ('command-r-plus',\n",
       "  'gpt-4-turbo-preview',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com',\n",
       "  'meta_llama3-70b-instruct-v1_0'),\n",
       " ('command-r-plus',\n",
       "  'gpt-4-turbo-preview',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com',\n",
       "  'mistral-large-latest'),\n",
       " ('command-r-plus',\n",
       "  'gpt-4-turbo-preview',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('command-r-plus',\n",
       "  'gpt-4-turbo-preview',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com',\n",
       "  'meta_llama3-70b-instruct-v1_0'),\n",
       " ('command-r-plus',\n",
       "  'gpt-4-turbo-preview',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com',\n",
       "  'mistral-large-latest'),\n",
       " ('command-r-plus',\n",
       "  'gpt-4-turbo-preview',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('command-r-plus',\n",
       "  'gpt-4-turbo-preview',\n",
       "  'meta_llama3-70b-instruct-v1_0',\n",
       "  'mistral-large-latest'),\n",
       " ('command-r-plus',\n",
       "  'gpt-4-turbo-preview',\n",
       "  'meta_llama3-70b-instruct-v1_0',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('command-r-plus',\n",
       "  'gpt-4-turbo-preview',\n",
       "  'mistral-large-latest',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('command-r-plus',\n",
       "  'gpt-4o-mini-2024-07-18',\n",
       "  'gpt-4o',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com'),\n",
       " ('command-r-plus',\n",
       "  'gpt-4o-mini-2024-07-18',\n",
       "  'gpt-4o',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com'),\n",
       " ('command-r-plus',\n",
       "  'gpt-4o-mini-2024-07-18',\n",
       "  'gpt-4o',\n",
       "  'meta_llama3-70b-instruct-v1_0'),\n",
       " ('command-r-plus',\n",
       "  'gpt-4o-mini-2024-07-18',\n",
       "  'gpt-4o',\n",
       "  'mistral-large-latest'),\n",
       " ('command-r-plus', 'gpt-4o-mini-2024-07-18', 'gpt-4o', 'open-mixtral-8x22b'),\n",
       " ('command-r-plus',\n",
       "  'gpt-4o-mini-2024-07-18',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com'),\n",
       " ('command-r-plus',\n",
       "  'gpt-4o-mini-2024-07-18',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com',\n",
       "  'meta_llama3-70b-instruct-v1_0'),\n",
       " ('command-r-plus',\n",
       "  'gpt-4o-mini-2024-07-18',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com',\n",
       "  'mistral-large-latest'),\n",
       " ('command-r-plus',\n",
       "  'gpt-4o-mini-2024-07-18',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('command-r-plus',\n",
       "  'gpt-4o-mini-2024-07-18',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com',\n",
       "  'meta_llama3-70b-instruct-v1_0'),\n",
       " ('command-r-plus',\n",
       "  'gpt-4o-mini-2024-07-18',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com',\n",
       "  'mistral-large-latest'),\n",
       " ('command-r-plus',\n",
       "  'gpt-4o-mini-2024-07-18',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('command-r-plus',\n",
       "  'gpt-4o-mini-2024-07-18',\n",
       "  'meta_llama3-70b-instruct-v1_0',\n",
       "  'mistral-large-latest'),\n",
       " ('command-r-plus',\n",
       "  'gpt-4o-mini-2024-07-18',\n",
       "  'meta_llama3-70b-instruct-v1_0',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('command-r-plus',\n",
       "  'gpt-4o-mini-2024-07-18',\n",
       "  'mistral-large-latest',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('command-r-plus',\n",
       "  'gpt-4o',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com'),\n",
       " ('command-r-plus',\n",
       "  'gpt-4o',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com',\n",
       "  'meta_llama3-70b-instruct-v1_0'),\n",
       " ('command-r-plus',\n",
       "  'gpt-4o',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com',\n",
       "  'mistral-large-latest'),\n",
       " ('command-r-plus',\n",
       "  'gpt-4o',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('command-r-plus',\n",
       "  'gpt-4o',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com',\n",
       "  'meta_llama3-70b-instruct-v1_0'),\n",
       " ('command-r-plus',\n",
       "  'gpt-4o',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com',\n",
       "  'mistral-large-latest'),\n",
       " ('command-r-plus',\n",
       "  'gpt-4o',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('command-r-plus',\n",
       "  'gpt-4o',\n",
       "  'meta_llama3-70b-instruct-v1_0',\n",
       "  'mistral-large-latest'),\n",
       " ('command-r-plus',\n",
       "  'gpt-4o',\n",
       "  'meta_llama3-70b-instruct-v1_0',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('command-r-plus', 'gpt-4o', 'mistral-large-latest', 'open-mixtral-8x22b'),\n",
       " ('command-r-plus',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com',\n",
       "  'meta_llama3-70b-instruct-v1_0'),\n",
       " ('command-r-plus',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com',\n",
       "  'mistral-large-latest'),\n",
       " ('command-r-plus',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('command-r-plus',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com',\n",
       "  'meta_llama3-70b-instruct-v1_0',\n",
       "  'mistral-large-latest'),\n",
       " ('command-r-plus',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com',\n",
       "  'meta_llama3-70b-instruct-v1_0',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('command-r-plus',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com',\n",
       "  'mistral-large-latest',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('command-r-plus',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com',\n",
       "  'meta_llama3-70b-instruct-v1_0',\n",
       "  'mistral-large-latest'),\n",
       " ('command-r-plus',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com',\n",
       "  'meta_llama3-70b-instruct-v1_0',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('command-r-plus',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com',\n",
       "  'mistral-large-latest',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('command-r-plus',\n",
       "  'meta_llama3-70b-instruct-v1_0',\n",
       "  'mistral-large-latest',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('gemini-1_5-pro', 'gpt-4-turbo-preview', 'gpt-4o-mini-2024-07-18', 'gpt-4o'),\n",
       " ('gemini-1_5-pro',\n",
       "  'gpt-4-turbo-preview',\n",
       "  'gpt-4o-mini-2024-07-18',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com'),\n",
       " ('gemini-1_5-pro',\n",
       "  'gpt-4-turbo-preview',\n",
       "  'gpt-4o-mini-2024-07-18',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com'),\n",
       " ('gemini-1_5-pro',\n",
       "  'gpt-4-turbo-preview',\n",
       "  'gpt-4o-mini-2024-07-18',\n",
       "  'meta_llama3-70b-instruct-v1_0'),\n",
       " ('gemini-1_5-pro',\n",
       "  'gpt-4-turbo-preview',\n",
       "  'gpt-4o-mini-2024-07-18',\n",
       "  'mistral-large-latest'),\n",
       " ('gemini-1_5-pro',\n",
       "  'gpt-4-turbo-preview',\n",
       "  'gpt-4o-mini-2024-07-18',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('gemini-1_5-pro',\n",
       "  'gpt-4-turbo-preview',\n",
       "  'gpt-4o',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com'),\n",
       " ('gemini-1_5-pro',\n",
       "  'gpt-4-turbo-preview',\n",
       "  'gpt-4o',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com'),\n",
       " ('gemini-1_5-pro',\n",
       "  'gpt-4-turbo-preview',\n",
       "  'gpt-4o',\n",
       "  'meta_llama3-70b-instruct-v1_0'),\n",
       " ('gemini-1_5-pro', 'gpt-4-turbo-preview', 'gpt-4o', 'mistral-large-latest'),\n",
       " ('gemini-1_5-pro', 'gpt-4-turbo-preview', 'gpt-4o', 'open-mixtral-8x22b'),\n",
       " ('gemini-1_5-pro',\n",
       "  'gpt-4-turbo-preview',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com'),\n",
       " ('gemini-1_5-pro',\n",
       "  'gpt-4-turbo-preview',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com',\n",
       "  'meta_llama3-70b-instruct-v1_0'),\n",
       " ('gemini-1_5-pro',\n",
       "  'gpt-4-turbo-preview',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com',\n",
       "  'mistral-large-latest'),\n",
       " ('gemini-1_5-pro',\n",
       "  'gpt-4-turbo-preview',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('gemini-1_5-pro',\n",
       "  'gpt-4-turbo-preview',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com',\n",
       "  'meta_llama3-70b-instruct-v1_0'),\n",
       " ('gemini-1_5-pro',\n",
       "  'gpt-4-turbo-preview',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com',\n",
       "  'mistral-large-latest'),\n",
       " ('gemini-1_5-pro',\n",
       "  'gpt-4-turbo-preview',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('gemini-1_5-pro',\n",
       "  'gpt-4-turbo-preview',\n",
       "  'meta_llama3-70b-instruct-v1_0',\n",
       "  'mistral-large-latest'),\n",
       " ('gemini-1_5-pro',\n",
       "  'gpt-4-turbo-preview',\n",
       "  'meta_llama3-70b-instruct-v1_0',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('gemini-1_5-pro',\n",
       "  'gpt-4-turbo-preview',\n",
       "  'mistral-large-latest',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('gemini-1_5-pro',\n",
       "  'gpt-4o-mini-2024-07-18',\n",
       "  'gpt-4o',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com'),\n",
       " ('gemini-1_5-pro',\n",
       "  'gpt-4o-mini-2024-07-18',\n",
       "  'gpt-4o',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com'),\n",
       " ('gemini-1_5-pro',\n",
       "  'gpt-4o-mini-2024-07-18',\n",
       "  'gpt-4o',\n",
       "  'meta_llama3-70b-instruct-v1_0'),\n",
       " ('gemini-1_5-pro',\n",
       "  'gpt-4o-mini-2024-07-18',\n",
       "  'gpt-4o',\n",
       "  'mistral-large-latest'),\n",
       " ('gemini-1_5-pro', 'gpt-4o-mini-2024-07-18', 'gpt-4o', 'open-mixtral-8x22b'),\n",
       " ('gemini-1_5-pro',\n",
       "  'gpt-4o-mini-2024-07-18',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com'),\n",
       " ('gemini-1_5-pro',\n",
       "  'gpt-4o-mini-2024-07-18',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com',\n",
       "  'meta_llama3-70b-instruct-v1_0'),\n",
       " ('gemini-1_5-pro',\n",
       "  'gpt-4o-mini-2024-07-18',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com',\n",
       "  'mistral-large-latest'),\n",
       " ('gemini-1_5-pro',\n",
       "  'gpt-4o-mini-2024-07-18',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('gemini-1_5-pro',\n",
       "  'gpt-4o-mini-2024-07-18',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com',\n",
       "  'meta_llama3-70b-instruct-v1_0'),\n",
       " ('gemini-1_5-pro',\n",
       "  'gpt-4o-mini-2024-07-18',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com',\n",
       "  'mistral-large-latest'),\n",
       " ('gemini-1_5-pro',\n",
       "  'gpt-4o-mini-2024-07-18',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('gemini-1_5-pro',\n",
       "  'gpt-4o-mini-2024-07-18',\n",
       "  'meta_llama3-70b-instruct-v1_0',\n",
       "  'mistral-large-latest'),\n",
       " ('gemini-1_5-pro',\n",
       "  'gpt-4o-mini-2024-07-18',\n",
       "  'meta_llama3-70b-instruct-v1_0',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('gemini-1_5-pro',\n",
       "  'gpt-4o-mini-2024-07-18',\n",
       "  'mistral-large-latest',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('gemini-1_5-pro',\n",
       "  'gpt-4o',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com'),\n",
       " ('gemini-1_5-pro',\n",
       "  'gpt-4o',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com',\n",
       "  'meta_llama3-70b-instruct-v1_0'),\n",
       " ('gemini-1_5-pro',\n",
       "  'gpt-4o',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com',\n",
       "  'mistral-large-latest'),\n",
       " ('gemini-1_5-pro',\n",
       "  'gpt-4o',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('gemini-1_5-pro',\n",
       "  'gpt-4o',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com',\n",
       "  'meta_llama3-70b-instruct-v1_0'),\n",
       " ('gemini-1_5-pro',\n",
       "  'gpt-4o',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com',\n",
       "  'mistral-large-latest'),\n",
       " ('gemini-1_5-pro',\n",
       "  'gpt-4o',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('gemini-1_5-pro',\n",
       "  'gpt-4o',\n",
       "  'meta_llama3-70b-instruct-v1_0',\n",
       "  'mistral-large-latest'),\n",
       " ('gemini-1_5-pro',\n",
       "  'gpt-4o',\n",
       "  'meta_llama3-70b-instruct-v1_0',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('gemini-1_5-pro', 'gpt-4o', 'mistral-large-latest', 'open-mixtral-8x22b'),\n",
       " ('gemini-1_5-pro',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com',\n",
       "  'meta_llama3-70b-instruct-v1_0'),\n",
       " ('gemini-1_5-pro',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com',\n",
       "  'mistral-large-latest'),\n",
       " ('gemini-1_5-pro',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('gemini-1_5-pro',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com',\n",
       "  'meta_llama3-70b-instruct-v1_0',\n",
       "  'mistral-large-latest'),\n",
       " ('gemini-1_5-pro',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com',\n",
       "  'meta_llama3-70b-instruct-v1_0',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('gemini-1_5-pro',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com',\n",
       "  'mistral-large-latest',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('gemini-1_5-pro',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com',\n",
       "  'meta_llama3-70b-instruct-v1_0',\n",
       "  'mistral-large-latest'),\n",
       " ('gemini-1_5-pro',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com',\n",
       "  'meta_llama3-70b-instruct-v1_0',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('gemini-1_5-pro',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com',\n",
       "  'mistral-large-latest',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('gemini-1_5-pro',\n",
       "  'meta_llama3-70b-instruct-v1_0',\n",
       "  'mistral-large-latest',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('gpt-4-turbo-preview',\n",
       "  'gpt-4o-mini-2024-07-18',\n",
       "  'gpt-4o',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com'),\n",
       " ('gpt-4-turbo-preview',\n",
       "  'gpt-4o-mini-2024-07-18',\n",
       "  'gpt-4o',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com'),\n",
       " ('gpt-4-turbo-preview',\n",
       "  'gpt-4o-mini-2024-07-18',\n",
       "  'gpt-4o',\n",
       "  'meta_llama3-70b-instruct-v1_0'),\n",
       " ('gpt-4-turbo-preview',\n",
       "  'gpt-4o-mini-2024-07-18',\n",
       "  'gpt-4o',\n",
       "  'mistral-large-latest'),\n",
       " ('gpt-4-turbo-preview',\n",
       "  'gpt-4o-mini-2024-07-18',\n",
       "  'gpt-4o',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('gpt-4-turbo-preview',\n",
       "  'gpt-4o-mini-2024-07-18',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com'),\n",
       " ('gpt-4-turbo-preview',\n",
       "  'gpt-4o-mini-2024-07-18',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com',\n",
       "  'meta_llama3-70b-instruct-v1_0'),\n",
       " ('gpt-4-turbo-preview',\n",
       "  'gpt-4o-mini-2024-07-18',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com',\n",
       "  'mistral-large-latest'),\n",
       " ('gpt-4-turbo-preview',\n",
       "  'gpt-4o-mini-2024-07-18',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('gpt-4-turbo-preview',\n",
       "  'gpt-4o-mini-2024-07-18',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com',\n",
       "  'meta_llama3-70b-instruct-v1_0'),\n",
       " ('gpt-4-turbo-preview',\n",
       "  'gpt-4o-mini-2024-07-18',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com',\n",
       "  'mistral-large-latest'),\n",
       " ('gpt-4-turbo-preview',\n",
       "  'gpt-4o-mini-2024-07-18',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('gpt-4-turbo-preview',\n",
       "  'gpt-4o-mini-2024-07-18',\n",
       "  'meta_llama3-70b-instruct-v1_0',\n",
       "  'mistral-large-latest'),\n",
       " ('gpt-4-turbo-preview',\n",
       "  'gpt-4o-mini-2024-07-18',\n",
       "  'meta_llama3-70b-instruct-v1_0',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('gpt-4-turbo-preview',\n",
       "  'gpt-4o-mini-2024-07-18',\n",
       "  'mistral-large-latest',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('gpt-4-turbo-preview',\n",
       "  'gpt-4o',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com'),\n",
       " ('gpt-4-turbo-preview',\n",
       "  'gpt-4o',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com',\n",
       "  'meta_llama3-70b-instruct-v1_0'),\n",
       " ('gpt-4-turbo-preview',\n",
       "  'gpt-4o',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com',\n",
       "  'mistral-large-latest'),\n",
       " ('gpt-4-turbo-preview',\n",
       "  'gpt-4o',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('gpt-4-turbo-preview',\n",
       "  'gpt-4o',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com',\n",
       "  'meta_llama3-70b-instruct-v1_0'),\n",
       " ('gpt-4-turbo-preview',\n",
       "  'gpt-4o',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com',\n",
       "  'mistral-large-latest'),\n",
       " ('gpt-4-turbo-preview',\n",
       "  'gpt-4o',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('gpt-4-turbo-preview',\n",
       "  'gpt-4o',\n",
       "  'meta_llama3-70b-instruct-v1_0',\n",
       "  'mistral-large-latest'),\n",
       " ('gpt-4-turbo-preview',\n",
       "  'gpt-4o',\n",
       "  'meta_llama3-70b-instruct-v1_0',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('gpt-4-turbo-preview',\n",
       "  'gpt-4o',\n",
       "  'mistral-large-latest',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('gpt-4-turbo-preview',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com',\n",
       "  'meta_llama3-70b-instruct-v1_0'),\n",
       " ('gpt-4-turbo-preview',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com',\n",
       "  'mistral-large-latest'),\n",
       " ('gpt-4-turbo-preview',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('gpt-4-turbo-preview',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com',\n",
       "  'meta_llama3-70b-instruct-v1_0',\n",
       "  'mistral-large-latest'),\n",
       " ('gpt-4-turbo-preview',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com',\n",
       "  'meta_llama3-70b-instruct-v1_0',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('gpt-4-turbo-preview',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com',\n",
       "  'mistral-large-latest',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('gpt-4-turbo-preview',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com',\n",
       "  'meta_llama3-70b-instruct-v1_0',\n",
       "  'mistral-large-latest'),\n",
       " ('gpt-4-turbo-preview',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com',\n",
       "  'meta_llama3-70b-instruct-v1_0',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('gpt-4-turbo-preview',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com',\n",
       "  'mistral-large-latest',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('gpt-4-turbo-preview',\n",
       "  'meta_llama3-70b-instruct-v1_0',\n",
       "  'mistral-large-latest',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('gpt-4o-mini-2024-07-18',\n",
       "  'gpt-4o',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com'),\n",
       " ('gpt-4o-mini-2024-07-18',\n",
       "  'gpt-4o',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com',\n",
       "  'meta_llama3-70b-instruct-v1_0'),\n",
       " ('gpt-4o-mini-2024-07-18',\n",
       "  'gpt-4o',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com',\n",
       "  'mistral-large-latest'),\n",
       " ('gpt-4o-mini-2024-07-18',\n",
       "  'gpt-4o',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('gpt-4o-mini-2024-07-18',\n",
       "  'gpt-4o',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com',\n",
       "  'meta_llama3-70b-instruct-v1_0'),\n",
       " ('gpt-4o-mini-2024-07-18',\n",
       "  'gpt-4o',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com',\n",
       "  'mistral-large-latest'),\n",
       " ('gpt-4o-mini-2024-07-18',\n",
       "  'gpt-4o',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('gpt-4o-mini-2024-07-18',\n",
       "  'gpt-4o',\n",
       "  'meta_llama3-70b-instruct-v1_0',\n",
       "  'mistral-large-latest'),\n",
       " ('gpt-4o-mini-2024-07-18',\n",
       "  'gpt-4o',\n",
       "  'meta_llama3-70b-instruct-v1_0',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('gpt-4o-mini-2024-07-18',\n",
       "  'gpt-4o',\n",
       "  'mistral-large-latest',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('gpt-4o-mini-2024-07-18',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com',\n",
       "  'meta_llama3-70b-instruct-v1_0'),\n",
       " ('gpt-4o-mini-2024-07-18',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com',\n",
       "  'mistral-large-latest'),\n",
       " ('gpt-4o-mini-2024-07-18',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('gpt-4o-mini-2024-07-18',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com',\n",
       "  'meta_llama3-70b-instruct-v1_0',\n",
       "  'mistral-large-latest'),\n",
       " ('gpt-4o-mini-2024-07-18',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com',\n",
       "  'meta_llama3-70b-instruct-v1_0',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('gpt-4o-mini-2024-07-18',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com',\n",
       "  'mistral-large-latest',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('gpt-4o-mini-2024-07-18',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com',\n",
       "  'meta_llama3-70b-instruct-v1_0',\n",
       "  'mistral-large-latest'),\n",
       " ('gpt-4o-mini-2024-07-18',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com',\n",
       "  'meta_llama3-70b-instruct-v1_0',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('gpt-4o-mini-2024-07-18',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com',\n",
       "  'mistral-large-latest',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('gpt-4o-mini-2024-07-18',\n",
       "  'meta_llama3-70b-instruct-v1_0',\n",
       "  'mistral-large-latest',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('gpt-4o',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com',\n",
       "  'meta_llama3-70b-instruct-v1_0'),\n",
       " ('gpt-4o',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com',\n",
       "  'mistral-large-latest'),\n",
       " ('gpt-4o',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('gpt-4o',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com',\n",
       "  'meta_llama3-70b-instruct-v1_0',\n",
       "  'mistral-large-latest'),\n",
       " ('gpt-4o',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com',\n",
       "  'meta_llama3-70b-instruct-v1_0',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('gpt-4o',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com',\n",
       "  'mistral-large-latest',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('gpt-4o',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com',\n",
       "  'meta_llama3-70b-instruct-v1_0',\n",
       "  'mistral-large-latest'),\n",
       " ('gpt-4o',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com',\n",
       "  'meta_llama3-70b-instruct-v1_0',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('gpt-4o',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com',\n",
       "  'mistral-large-latest',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('gpt-4o',\n",
       "  'meta_llama3-70b-instruct-v1_0',\n",
       "  'mistral-large-latest',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com',\n",
       "  'meta_llama3-70b-instruct-v1_0',\n",
       "  'mistral-large-latest'),\n",
       " ('Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com',\n",
       "  'meta_llama3-70b-instruct-v1_0',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com',\n",
       "  'mistral-large-latest',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com',\n",
       "  'meta_llama3-70b-instruct-v1_0',\n",
       "  'mistral-large-latest',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com',\n",
       "  'meta_llama3-70b-instruct-v1_0',\n",
       "  'mistral-large-latest',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'claude-3-opus-20240229',\n",
       "  'command-r-plus',\n",
       "  'gemini-1_5-pro',\n",
       "  'gpt-4-turbo-preview'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'claude-3-opus-20240229',\n",
       "  'command-r-plus',\n",
       "  'gemini-1_5-pro',\n",
       "  'gpt-4o-mini-2024-07-18'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'claude-3-opus-20240229',\n",
       "  'command-r-plus',\n",
       "  'gemini-1_5-pro',\n",
       "  'gpt-4o'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'claude-3-opus-20240229',\n",
       "  'command-r-plus',\n",
       "  'gemini-1_5-pro',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'claude-3-opus-20240229',\n",
       "  'command-r-plus',\n",
       "  'gemini-1_5-pro',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'claude-3-opus-20240229',\n",
       "  'command-r-plus',\n",
       "  'gemini-1_5-pro',\n",
       "  'meta_llama3-70b-instruct-v1_0'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'claude-3-opus-20240229',\n",
       "  'command-r-plus',\n",
       "  'gemini-1_5-pro',\n",
       "  'mistral-large-latest'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'claude-3-opus-20240229',\n",
       "  'command-r-plus',\n",
       "  'gemini-1_5-pro',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'claude-3-opus-20240229',\n",
       "  'command-r-plus',\n",
       "  'gpt-4-turbo-preview',\n",
       "  'gpt-4o-mini-2024-07-18'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'claude-3-opus-20240229',\n",
       "  'command-r-plus',\n",
       "  'gpt-4-turbo-preview',\n",
       "  'gpt-4o'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'claude-3-opus-20240229',\n",
       "  'command-r-plus',\n",
       "  'gpt-4-turbo-preview',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'claude-3-opus-20240229',\n",
       "  'command-r-plus',\n",
       "  'gpt-4-turbo-preview',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'claude-3-opus-20240229',\n",
       "  'command-r-plus',\n",
       "  'gpt-4-turbo-preview',\n",
       "  'meta_llama3-70b-instruct-v1_0'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'claude-3-opus-20240229',\n",
       "  'command-r-plus',\n",
       "  'gpt-4-turbo-preview',\n",
       "  'mistral-large-latest'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'claude-3-opus-20240229',\n",
       "  'command-r-plus',\n",
       "  'gpt-4-turbo-preview',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'claude-3-opus-20240229',\n",
       "  'command-r-plus',\n",
       "  'gpt-4o-mini-2024-07-18',\n",
       "  'gpt-4o'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'claude-3-opus-20240229',\n",
       "  'command-r-plus',\n",
       "  'gpt-4o-mini-2024-07-18',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'claude-3-opus-20240229',\n",
       "  'command-r-plus',\n",
       "  'gpt-4o-mini-2024-07-18',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'claude-3-opus-20240229',\n",
       "  'command-r-plus',\n",
       "  'gpt-4o-mini-2024-07-18',\n",
       "  'meta_llama3-70b-instruct-v1_0'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'claude-3-opus-20240229',\n",
       "  'command-r-plus',\n",
       "  'gpt-4o-mini-2024-07-18',\n",
       "  'mistral-large-latest'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'claude-3-opus-20240229',\n",
       "  'command-r-plus',\n",
       "  'gpt-4o-mini-2024-07-18',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'claude-3-opus-20240229',\n",
       "  'command-r-plus',\n",
       "  'gpt-4o',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'claude-3-opus-20240229',\n",
       "  'command-r-plus',\n",
       "  'gpt-4o',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'claude-3-opus-20240229',\n",
       "  'command-r-plus',\n",
       "  'gpt-4o',\n",
       "  'meta_llama3-70b-instruct-v1_0'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'claude-3-opus-20240229',\n",
       "  'command-r-plus',\n",
       "  'gpt-4o',\n",
       "  'mistral-large-latest'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'claude-3-opus-20240229',\n",
       "  'command-r-plus',\n",
       "  'gpt-4o',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'claude-3-opus-20240229',\n",
       "  'command-r-plus',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'claude-3-opus-20240229',\n",
       "  'command-r-plus',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com',\n",
       "  'meta_llama3-70b-instruct-v1_0'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'claude-3-opus-20240229',\n",
       "  'command-r-plus',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com',\n",
       "  'mistral-large-latest'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'claude-3-opus-20240229',\n",
       "  'command-r-plus',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'claude-3-opus-20240229',\n",
       "  'command-r-plus',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com',\n",
       "  'meta_llama3-70b-instruct-v1_0'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'claude-3-opus-20240229',\n",
       "  'command-r-plus',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com',\n",
       "  'mistral-large-latest'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'claude-3-opus-20240229',\n",
       "  'command-r-plus',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'claude-3-opus-20240229',\n",
       "  'command-r-plus',\n",
       "  'meta_llama3-70b-instruct-v1_0',\n",
       "  'mistral-large-latest'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'claude-3-opus-20240229',\n",
       "  'command-r-plus',\n",
       "  'meta_llama3-70b-instruct-v1_0',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'claude-3-opus-20240229',\n",
       "  'command-r-plus',\n",
       "  'mistral-large-latest',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'claude-3-opus-20240229',\n",
       "  'gemini-1_5-pro',\n",
       "  'gpt-4-turbo-preview',\n",
       "  'gpt-4o-mini-2024-07-18'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'claude-3-opus-20240229',\n",
       "  'gemini-1_5-pro',\n",
       "  'gpt-4-turbo-preview',\n",
       "  'gpt-4o'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'claude-3-opus-20240229',\n",
       "  'gemini-1_5-pro',\n",
       "  'gpt-4-turbo-preview',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'claude-3-opus-20240229',\n",
       "  'gemini-1_5-pro',\n",
       "  'gpt-4-turbo-preview',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'claude-3-opus-20240229',\n",
       "  'gemini-1_5-pro',\n",
       "  'gpt-4-turbo-preview',\n",
       "  'meta_llama3-70b-instruct-v1_0'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'claude-3-opus-20240229',\n",
       "  'gemini-1_5-pro',\n",
       "  'gpt-4-turbo-preview',\n",
       "  'mistral-large-latest'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'claude-3-opus-20240229',\n",
       "  'gemini-1_5-pro',\n",
       "  'gpt-4-turbo-preview',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'claude-3-opus-20240229',\n",
       "  'gemini-1_5-pro',\n",
       "  'gpt-4o-mini-2024-07-18',\n",
       "  'gpt-4o'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'claude-3-opus-20240229',\n",
       "  'gemini-1_5-pro',\n",
       "  'gpt-4o-mini-2024-07-18',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'claude-3-opus-20240229',\n",
       "  'gemini-1_5-pro',\n",
       "  'gpt-4o-mini-2024-07-18',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'claude-3-opus-20240229',\n",
       "  'gemini-1_5-pro',\n",
       "  'gpt-4o-mini-2024-07-18',\n",
       "  'meta_llama3-70b-instruct-v1_0'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'claude-3-opus-20240229',\n",
       "  'gemini-1_5-pro',\n",
       "  'gpt-4o-mini-2024-07-18',\n",
       "  'mistral-large-latest'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'claude-3-opus-20240229',\n",
       "  'gemini-1_5-pro',\n",
       "  'gpt-4o-mini-2024-07-18',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'claude-3-opus-20240229',\n",
       "  'gemini-1_5-pro',\n",
       "  'gpt-4o',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'claude-3-opus-20240229',\n",
       "  'gemini-1_5-pro',\n",
       "  'gpt-4o',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'claude-3-opus-20240229',\n",
       "  'gemini-1_5-pro',\n",
       "  'gpt-4o',\n",
       "  'meta_llama3-70b-instruct-v1_0'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'claude-3-opus-20240229',\n",
       "  'gemini-1_5-pro',\n",
       "  'gpt-4o',\n",
       "  'mistral-large-latest'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'claude-3-opus-20240229',\n",
       "  'gemini-1_5-pro',\n",
       "  'gpt-4o',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'claude-3-opus-20240229',\n",
       "  'gemini-1_5-pro',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'claude-3-opus-20240229',\n",
       "  'gemini-1_5-pro',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com',\n",
       "  'meta_llama3-70b-instruct-v1_0'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'claude-3-opus-20240229',\n",
       "  'gemini-1_5-pro',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com',\n",
       "  'mistral-large-latest'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'claude-3-opus-20240229',\n",
       "  'gemini-1_5-pro',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'claude-3-opus-20240229',\n",
       "  'gemini-1_5-pro',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com',\n",
       "  'meta_llama3-70b-instruct-v1_0'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'claude-3-opus-20240229',\n",
       "  'gemini-1_5-pro',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com',\n",
       "  'mistral-large-latest'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'claude-3-opus-20240229',\n",
       "  'gemini-1_5-pro',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'claude-3-opus-20240229',\n",
       "  'gemini-1_5-pro',\n",
       "  'meta_llama3-70b-instruct-v1_0',\n",
       "  'mistral-large-latest'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'claude-3-opus-20240229',\n",
       "  'gemini-1_5-pro',\n",
       "  'meta_llama3-70b-instruct-v1_0',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'claude-3-opus-20240229',\n",
       "  'gemini-1_5-pro',\n",
       "  'mistral-large-latest',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'claude-3-opus-20240229',\n",
       "  'gpt-4-turbo-preview',\n",
       "  'gpt-4o-mini-2024-07-18',\n",
       "  'gpt-4o'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'claude-3-opus-20240229',\n",
       "  'gpt-4-turbo-preview',\n",
       "  'gpt-4o-mini-2024-07-18',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'claude-3-opus-20240229',\n",
       "  'gpt-4-turbo-preview',\n",
       "  'gpt-4o-mini-2024-07-18',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'claude-3-opus-20240229',\n",
       "  'gpt-4-turbo-preview',\n",
       "  'gpt-4o-mini-2024-07-18',\n",
       "  'meta_llama3-70b-instruct-v1_0'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'claude-3-opus-20240229',\n",
       "  'gpt-4-turbo-preview',\n",
       "  'gpt-4o-mini-2024-07-18',\n",
       "  'mistral-large-latest'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'claude-3-opus-20240229',\n",
       "  'gpt-4-turbo-preview',\n",
       "  'gpt-4o-mini-2024-07-18',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'claude-3-opus-20240229',\n",
       "  'gpt-4-turbo-preview',\n",
       "  'gpt-4o',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'claude-3-opus-20240229',\n",
       "  'gpt-4-turbo-preview',\n",
       "  'gpt-4o',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'claude-3-opus-20240229',\n",
       "  'gpt-4-turbo-preview',\n",
       "  'gpt-4o',\n",
       "  'meta_llama3-70b-instruct-v1_0'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'claude-3-opus-20240229',\n",
       "  'gpt-4-turbo-preview',\n",
       "  'gpt-4o',\n",
       "  'mistral-large-latest'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'claude-3-opus-20240229',\n",
       "  'gpt-4-turbo-preview',\n",
       "  'gpt-4o',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'claude-3-opus-20240229',\n",
       "  'gpt-4-turbo-preview',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'claude-3-opus-20240229',\n",
       "  'gpt-4-turbo-preview',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com',\n",
       "  'meta_llama3-70b-instruct-v1_0'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'claude-3-opus-20240229',\n",
       "  'gpt-4-turbo-preview',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com',\n",
       "  'mistral-large-latest'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'claude-3-opus-20240229',\n",
       "  'gpt-4-turbo-preview',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'claude-3-opus-20240229',\n",
       "  'gpt-4-turbo-preview',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com',\n",
       "  'meta_llama3-70b-instruct-v1_0'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'claude-3-opus-20240229',\n",
       "  'gpt-4-turbo-preview',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com',\n",
       "  'mistral-large-latest'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'claude-3-opus-20240229',\n",
       "  'gpt-4-turbo-preview',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'claude-3-opus-20240229',\n",
       "  'gpt-4-turbo-preview',\n",
       "  'meta_llama3-70b-instruct-v1_0',\n",
       "  'mistral-large-latest'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'claude-3-opus-20240229',\n",
       "  'gpt-4-turbo-preview',\n",
       "  'meta_llama3-70b-instruct-v1_0',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'claude-3-opus-20240229',\n",
       "  'gpt-4-turbo-preview',\n",
       "  'mistral-large-latest',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'claude-3-opus-20240229',\n",
       "  'gpt-4o-mini-2024-07-18',\n",
       "  'gpt-4o',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'claude-3-opus-20240229',\n",
       "  'gpt-4o-mini-2024-07-18',\n",
       "  'gpt-4o',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'claude-3-opus-20240229',\n",
       "  'gpt-4o-mini-2024-07-18',\n",
       "  'gpt-4o',\n",
       "  'meta_llama3-70b-instruct-v1_0'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'claude-3-opus-20240229',\n",
       "  'gpt-4o-mini-2024-07-18',\n",
       "  'gpt-4o',\n",
       "  'mistral-large-latest'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'claude-3-opus-20240229',\n",
       "  'gpt-4o-mini-2024-07-18',\n",
       "  'gpt-4o',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'claude-3-opus-20240229',\n",
       "  'gpt-4o-mini-2024-07-18',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'claude-3-opus-20240229',\n",
       "  'gpt-4o-mini-2024-07-18',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com',\n",
       "  'meta_llama3-70b-instruct-v1_0'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'claude-3-opus-20240229',\n",
       "  'gpt-4o-mini-2024-07-18',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com',\n",
       "  'mistral-large-latest'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'claude-3-opus-20240229',\n",
       "  'gpt-4o-mini-2024-07-18',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'claude-3-opus-20240229',\n",
       "  'gpt-4o-mini-2024-07-18',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com',\n",
       "  'meta_llama3-70b-instruct-v1_0'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'claude-3-opus-20240229',\n",
       "  'gpt-4o-mini-2024-07-18',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com',\n",
       "  'mistral-large-latest'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'claude-3-opus-20240229',\n",
       "  'gpt-4o-mini-2024-07-18',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'claude-3-opus-20240229',\n",
       "  'gpt-4o-mini-2024-07-18',\n",
       "  'meta_llama3-70b-instruct-v1_0',\n",
       "  'mistral-large-latest'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'claude-3-opus-20240229',\n",
       "  'gpt-4o-mini-2024-07-18',\n",
       "  'meta_llama3-70b-instruct-v1_0',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'claude-3-opus-20240229',\n",
       "  'gpt-4o-mini-2024-07-18',\n",
       "  'mistral-large-latest',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'claude-3-opus-20240229',\n",
       "  'gpt-4o',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'claude-3-opus-20240229',\n",
       "  'gpt-4o',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com',\n",
       "  'meta_llama3-70b-instruct-v1_0'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'claude-3-opus-20240229',\n",
       "  'gpt-4o',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com',\n",
       "  'mistral-large-latest'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'claude-3-opus-20240229',\n",
       "  'gpt-4o',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'claude-3-opus-20240229',\n",
       "  'gpt-4o',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com',\n",
       "  'meta_llama3-70b-instruct-v1_0'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'claude-3-opus-20240229',\n",
       "  'gpt-4o',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com',\n",
       "  'mistral-large-latest'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'claude-3-opus-20240229',\n",
       "  'gpt-4o',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'claude-3-opus-20240229',\n",
       "  'gpt-4o',\n",
       "  'meta_llama3-70b-instruct-v1_0',\n",
       "  'mistral-large-latest'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'claude-3-opus-20240229',\n",
       "  'gpt-4o',\n",
       "  'meta_llama3-70b-instruct-v1_0',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'claude-3-opus-20240229',\n",
       "  'gpt-4o',\n",
       "  'mistral-large-latest',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'claude-3-opus-20240229',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com',\n",
       "  'meta_llama3-70b-instruct-v1_0'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'claude-3-opus-20240229',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com',\n",
       "  'mistral-large-latest'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'claude-3-opus-20240229',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'claude-3-opus-20240229',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com',\n",
       "  'meta_llama3-70b-instruct-v1_0',\n",
       "  'mistral-large-latest'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'claude-3-opus-20240229',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com',\n",
       "  'meta_llama3-70b-instruct-v1_0',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'claude-3-opus-20240229',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com',\n",
       "  'mistral-large-latest',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'claude-3-opus-20240229',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com',\n",
       "  'meta_llama3-70b-instruct-v1_0',\n",
       "  'mistral-large-latest'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'claude-3-opus-20240229',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com',\n",
       "  'meta_llama3-70b-instruct-v1_0',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'claude-3-opus-20240229',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com',\n",
       "  'mistral-large-latest',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'claude-3-opus-20240229',\n",
       "  'meta_llama3-70b-instruct-v1_0',\n",
       "  'mistral-large-latest',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'command-r-plus',\n",
       "  'gemini-1_5-pro',\n",
       "  'gpt-4-turbo-preview',\n",
       "  'gpt-4o-mini-2024-07-18'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'command-r-plus',\n",
       "  'gemini-1_5-pro',\n",
       "  'gpt-4-turbo-preview',\n",
       "  'gpt-4o'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'command-r-plus',\n",
       "  'gemini-1_5-pro',\n",
       "  'gpt-4-turbo-preview',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'command-r-plus',\n",
       "  'gemini-1_5-pro',\n",
       "  'gpt-4-turbo-preview',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'command-r-plus',\n",
       "  'gemini-1_5-pro',\n",
       "  'gpt-4-turbo-preview',\n",
       "  'meta_llama3-70b-instruct-v1_0'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'command-r-plus',\n",
       "  'gemini-1_5-pro',\n",
       "  'gpt-4-turbo-preview',\n",
       "  'mistral-large-latest'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'command-r-plus',\n",
       "  'gemini-1_5-pro',\n",
       "  'gpt-4-turbo-preview',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'command-r-plus',\n",
       "  'gemini-1_5-pro',\n",
       "  'gpt-4o-mini-2024-07-18',\n",
       "  'gpt-4o'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'command-r-plus',\n",
       "  'gemini-1_5-pro',\n",
       "  'gpt-4o-mini-2024-07-18',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'command-r-plus',\n",
       "  'gemini-1_5-pro',\n",
       "  'gpt-4o-mini-2024-07-18',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'command-r-plus',\n",
       "  'gemini-1_5-pro',\n",
       "  'gpt-4o-mini-2024-07-18',\n",
       "  'meta_llama3-70b-instruct-v1_0'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'command-r-plus',\n",
       "  'gemini-1_5-pro',\n",
       "  'gpt-4o-mini-2024-07-18',\n",
       "  'mistral-large-latest'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'command-r-plus',\n",
       "  'gemini-1_5-pro',\n",
       "  'gpt-4o-mini-2024-07-18',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'command-r-plus',\n",
       "  'gemini-1_5-pro',\n",
       "  'gpt-4o',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'command-r-plus',\n",
       "  'gemini-1_5-pro',\n",
       "  'gpt-4o',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'command-r-plus',\n",
       "  'gemini-1_5-pro',\n",
       "  'gpt-4o',\n",
       "  'meta_llama3-70b-instruct-v1_0'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'command-r-plus',\n",
       "  'gemini-1_5-pro',\n",
       "  'gpt-4o',\n",
       "  'mistral-large-latest'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'command-r-plus',\n",
       "  'gemini-1_5-pro',\n",
       "  'gpt-4o',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'command-r-plus',\n",
       "  'gemini-1_5-pro',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'command-r-plus',\n",
       "  'gemini-1_5-pro',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com',\n",
       "  'meta_llama3-70b-instruct-v1_0'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'command-r-plus',\n",
       "  'gemini-1_5-pro',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com',\n",
       "  'mistral-large-latest'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'command-r-plus',\n",
       "  'gemini-1_5-pro',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'command-r-plus',\n",
       "  'gemini-1_5-pro',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com',\n",
       "  'meta_llama3-70b-instruct-v1_0'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'command-r-plus',\n",
       "  'gemini-1_5-pro',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com',\n",
       "  'mistral-large-latest'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'command-r-plus',\n",
       "  'gemini-1_5-pro',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'command-r-plus',\n",
       "  'gemini-1_5-pro',\n",
       "  'meta_llama3-70b-instruct-v1_0',\n",
       "  'mistral-large-latest'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'command-r-plus',\n",
       "  'gemini-1_5-pro',\n",
       "  'meta_llama3-70b-instruct-v1_0',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'command-r-plus',\n",
       "  'gemini-1_5-pro',\n",
       "  'mistral-large-latest',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'command-r-plus',\n",
       "  'gpt-4-turbo-preview',\n",
       "  'gpt-4o-mini-2024-07-18',\n",
       "  'gpt-4o'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'command-r-plus',\n",
       "  'gpt-4-turbo-preview',\n",
       "  'gpt-4o-mini-2024-07-18',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'command-r-plus',\n",
       "  'gpt-4-turbo-preview',\n",
       "  'gpt-4o-mini-2024-07-18',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'command-r-plus',\n",
       "  'gpt-4-turbo-preview',\n",
       "  'gpt-4o-mini-2024-07-18',\n",
       "  'meta_llama3-70b-instruct-v1_0'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'command-r-plus',\n",
       "  'gpt-4-turbo-preview',\n",
       "  'gpt-4o-mini-2024-07-18',\n",
       "  'mistral-large-latest'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'command-r-plus',\n",
       "  'gpt-4-turbo-preview',\n",
       "  'gpt-4o-mini-2024-07-18',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'command-r-plus',\n",
       "  'gpt-4-turbo-preview',\n",
       "  'gpt-4o',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'command-r-plus',\n",
       "  'gpt-4-turbo-preview',\n",
       "  'gpt-4o',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'command-r-plus',\n",
       "  'gpt-4-turbo-preview',\n",
       "  'gpt-4o',\n",
       "  'meta_llama3-70b-instruct-v1_0'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'command-r-plus',\n",
       "  'gpt-4-turbo-preview',\n",
       "  'gpt-4o',\n",
       "  'mistral-large-latest'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'command-r-plus',\n",
       "  'gpt-4-turbo-preview',\n",
       "  'gpt-4o',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'command-r-plus',\n",
       "  'gpt-4-turbo-preview',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'command-r-plus',\n",
       "  'gpt-4-turbo-preview',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com',\n",
       "  'meta_llama3-70b-instruct-v1_0'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'command-r-plus',\n",
       "  'gpt-4-turbo-preview',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com',\n",
       "  'mistral-large-latest'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'command-r-plus',\n",
       "  'gpt-4-turbo-preview',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'command-r-plus',\n",
       "  'gpt-4-turbo-preview',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com',\n",
       "  'meta_llama3-70b-instruct-v1_0'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'command-r-plus',\n",
       "  'gpt-4-turbo-preview',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com',\n",
       "  'mistral-large-latest'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'command-r-plus',\n",
       "  'gpt-4-turbo-preview',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'command-r-plus',\n",
       "  'gpt-4-turbo-preview',\n",
       "  'meta_llama3-70b-instruct-v1_0',\n",
       "  'mistral-large-latest'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'command-r-plus',\n",
       "  'gpt-4-turbo-preview',\n",
       "  'meta_llama3-70b-instruct-v1_0',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'command-r-plus',\n",
       "  'gpt-4-turbo-preview',\n",
       "  'mistral-large-latest',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'command-r-plus',\n",
       "  'gpt-4o-mini-2024-07-18',\n",
       "  'gpt-4o',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'command-r-plus',\n",
       "  'gpt-4o-mini-2024-07-18',\n",
       "  'gpt-4o',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'command-r-plus',\n",
       "  'gpt-4o-mini-2024-07-18',\n",
       "  'gpt-4o',\n",
       "  'meta_llama3-70b-instruct-v1_0'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'command-r-plus',\n",
       "  'gpt-4o-mini-2024-07-18',\n",
       "  'gpt-4o',\n",
       "  'mistral-large-latest'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'command-r-plus',\n",
       "  'gpt-4o-mini-2024-07-18',\n",
       "  'gpt-4o',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'command-r-plus',\n",
       "  'gpt-4o-mini-2024-07-18',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'command-r-plus',\n",
       "  'gpt-4o-mini-2024-07-18',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com',\n",
       "  'meta_llama3-70b-instruct-v1_0'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'command-r-plus',\n",
       "  'gpt-4o-mini-2024-07-18',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com',\n",
       "  'mistral-large-latest'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'command-r-plus',\n",
       "  'gpt-4o-mini-2024-07-18',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'command-r-plus',\n",
       "  'gpt-4o-mini-2024-07-18',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com',\n",
       "  'meta_llama3-70b-instruct-v1_0'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'command-r-plus',\n",
       "  'gpt-4o-mini-2024-07-18',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com',\n",
       "  'mistral-large-latest'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'command-r-plus',\n",
       "  'gpt-4o-mini-2024-07-18',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'command-r-plus',\n",
       "  'gpt-4o-mini-2024-07-18',\n",
       "  'meta_llama3-70b-instruct-v1_0',\n",
       "  'mistral-large-latest'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'command-r-plus',\n",
       "  'gpt-4o-mini-2024-07-18',\n",
       "  'meta_llama3-70b-instruct-v1_0',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'command-r-plus',\n",
       "  'gpt-4o-mini-2024-07-18',\n",
       "  'mistral-large-latest',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'command-r-plus',\n",
       "  'gpt-4o',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'command-r-plus',\n",
       "  'gpt-4o',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com',\n",
       "  'meta_llama3-70b-instruct-v1_0'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'command-r-plus',\n",
       "  'gpt-4o',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com',\n",
       "  'mistral-large-latest'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'command-r-plus',\n",
       "  'gpt-4o',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'command-r-plus',\n",
       "  'gpt-4o',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com',\n",
       "  'meta_llama3-70b-instruct-v1_0'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'command-r-plus',\n",
       "  'gpt-4o',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com',\n",
       "  'mistral-large-latest'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'command-r-plus',\n",
       "  'gpt-4o',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'command-r-plus',\n",
       "  'gpt-4o',\n",
       "  'meta_llama3-70b-instruct-v1_0',\n",
       "  'mistral-large-latest'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'command-r-plus',\n",
       "  'gpt-4o',\n",
       "  'meta_llama3-70b-instruct-v1_0',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'command-r-plus',\n",
       "  'gpt-4o',\n",
       "  'mistral-large-latest',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'command-r-plus',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com',\n",
       "  'meta_llama3-70b-instruct-v1_0'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'command-r-plus',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com',\n",
       "  'mistral-large-latest'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'command-r-plus',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'command-r-plus',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com',\n",
       "  'meta_llama3-70b-instruct-v1_0',\n",
       "  'mistral-large-latest'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'command-r-plus',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com',\n",
       "  'meta_llama3-70b-instruct-v1_0',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'command-r-plus',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com',\n",
       "  'mistral-large-latest',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'command-r-plus',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com',\n",
       "  'meta_llama3-70b-instruct-v1_0',\n",
       "  'mistral-large-latest'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'command-r-plus',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com',\n",
       "  'meta_llama3-70b-instruct-v1_0',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'command-r-plus',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com',\n",
       "  'mistral-large-latest',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'command-r-plus',\n",
       "  'meta_llama3-70b-instruct-v1_0',\n",
       "  'mistral-large-latest',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'gemini-1_5-pro',\n",
       "  'gpt-4-turbo-preview',\n",
       "  'gpt-4o-mini-2024-07-18',\n",
       "  'gpt-4o'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'gemini-1_5-pro',\n",
       "  'gpt-4-turbo-preview',\n",
       "  'gpt-4o-mini-2024-07-18',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'gemini-1_5-pro',\n",
       "  'gpt-4-turbo-preview',\n",
       "  'gpt-4o-mini-2024-07-18',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'gemini-1_5-pro',\n",
       "  'gpt-4-turbo-preview',\n",
       "  'gpt-4o-mini-2024-07-18',\n",
       "  'meta_llama3-70b-instruct-v1_0'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'gemini-1_5-pro',\n",
       "  'gpt-4-turbo-preview',\n",
       "  'gpt-4o-mini-2024-07-18',\n",
       "  'mistral-large-latest'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'gemini-1_5-pro',\n",
       "  'gpt-4-turbo-preview',\n",
       "  'gpt-4o-mini-2024-07-18',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'gemini-1_5-pro',\n",
       "  'gpt-4-turbo-preview',\n",
       "  'gpt-4o',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'gemini-1_5-pro',\n",
       "  'gpt-4-turbo-preview',\n",
       "  'gpt-4o',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'gemini-1_5-pro',\n",
       "  'gpt-4-turbo-preview',\n",
       "  'gpt-4o',\n",
       "  'meta_llama3-70b-instruct-v1_0'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'gemini-1_5-pro',\n",
       "  'gpt-4-turbo-preview',\n",
       "  'gpt-4o',\n",
       "  'mistral-large-latest'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'gemini-1_5-pro',\n",
       "  'gpt-4-turbo-preview',\n",
       "  'gpt-4o',\n",
       "  'open-mixtral-8x22b'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'gemini-1_5-pro',\n",
       "  'gpt-4-turbo-preview',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com',\n",
       "  'Meta-Llama-3-1-70B-Instruct-ostu_eastus_models_ai_azure_com'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'gemini-1_5-pro',\n",
       "  'gpt-4-turbo-preview',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com',\n",
       "  'meta_llama3-70b-instruct-v1_0'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'gemini-1_5-pro',\n",
       "  'gpt-4-turbo-preview',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com',\n",
       "  'mistral-large-latest'),\n",
       " ('claude-3-5-sonnet-20240620',\n",
       "  'gemini-1_5-pro',\n",
       "  'gpt-4-turbo-preview',\n",
       "  'Meta-Llama-3-1-405B-Instruct-jjo_eastus_models_ai_azure_com',\n",
       "  'open-mixtral-8x22b'),\n",
       " ...]"
      ]
     },
     "execution_count": 32,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from itertools import combinations\n",
    "\n",
    "def all_combinations(arr):\n",
    "    x = len(arr)\n",
    "    all_combos = []\n",
    "    for r in range(2, x + 1):\n",
    "        combos = list(combinations(arr, r))\n",
    "        all_combos.extend(combos)\n",
    "    return all_combos\n",
    "\n",
    "all_combos = all_combinations(llm_models)\n",
    "all_combos"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {},
   "outputs": [
    {
     "ename": "RuntimeError",
     "evalue": "No active exception to reraise",
     "output_type": "error",
     "traceback": [
      "\u001b[1;31m---------------------------------------------------------------------------\u001b[0m",
      "\u001b[1;31mRuntimeError\u001b[0m                              Traceback (most recent call last)",
      "Cell \u001b[1;32mIn[33], line 1\u001b[0m\n\u001b[1;32m----> 1\u001b[0m \u001b[38;5;28;01mraise\u001b[39;00m\n",
      "\u001b[1;31mRuntimeError\u001b[0m: No active exception to reraise"
     ]
    }
   ],
   "source": [
    "raise"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "  2%|▏         | 81/4083 [02:12<1:49:13,  1.64s/it]\n"
     ]
    },
    {
     "ename": "KeyboardInterrupt",
     "evalue": "",
     "output_type": "error",
     "traceback": [
      "\u001b[1;31m---------------------------------------------------------------------------\u001b[0m",
      "\u001b[1;31mKeyboardInterrupt\u001b[0m                         Traceback (most recent call last)",
      "Cell \u001b[1;32mIn[15], line 8\u001b[0m\n\u001b[0;32m      5\u001b[0m subset_norm_choice_cols \u001b[38;5;241m=\u001b[39m [col \u001b[38;5;28;01mfor\u001b[39;00m col \u001b[38;5;129;01min\u001b[39;00m norm_choice_cols \u001b[38;5;28;01mif\u001b[39;00m col[\u001b[38;5;241m0\u001b[39m] \u001b[38;5;129;01min\u001b[39;00m models]\n\u001b[0;32m      6\u001b[0m answer_values \u001b[38;5;241m=\u001b[39m moe_llm_answers_comb_df\u001b[38;5;241m.\u001b[39mapply(axis\u001b[38;5;241m=\u001b[39m\u001b[38;5;124m'\u001b[39m\u001b[38;5;124mcolumns\u001b[39m\u001b[38;5;124m'\u001b[39m, \n\u001b[0;32m      7\u001b[0m     func\u001b[38;5;241m=\u001b[39m\u001b[38;5;28;01mlambda\u001b[39;00m x: x[subset_norm_choice_cols]\u001b[38;5;241m.\u001b[39mvalue_counts())\n\u001b[1;32m----> 8\u001b[0m most_common_answer \u001b[38;5;241m=\u001b[39m \u001b[43mmoe_llm_answers_comb_df\u001b[49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mapply\u001b[49m\u001b[43m(\u001b[49m\u001b[43maxis\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[38;5;124;43m'\u001b[39;49m\u001b[38;5;124;43mcolumns\u001b[39;49m\u001b[38;5;124;43m'\u001b[39;49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\n\u001b[0;32m      9\u001b[0m \u001b[43m    \u001b[49m\u001b[43mfunc\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[38;5;28;43;01mlambda\u001b[39;49;00m\u001b[43m \u001b[49m\u001b[43mx\u001b[49m\u001b[43m:\u001b[49m\u001b[43m \u001b[49m\u001b[43mget_most_common_or_random\u001b[49m\u001b[43m(\u001b[49m\u001b[43mx\u001b[49m\u001b[43m[\u001b[49m\u001b[43msubset_norm_choice_cols\u001b[49m\u001b[43m]\u001b[49m\u001b[43m)\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43mresult_type\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[38;5;124;43m'\u001b[39;49m\u001b[38;5;124;43mexpand\u001b[39;49m\u001b[38;5;124;43m'\u001b[39;49m\u001b[43m)\u001b[49m\n\u001b[0;32m     10\u001b[0m largest_common_answer \u001b[38;5;241m=\u001b[39m moe_llm_answers_comb_df\u001b[38;5;241m.\u001b[39mapply(axis\u001b[38;5;241m=\u001b[39m\u001b[38;5;124m'\u001b[39m\u001b[38;5;124mcolumns\u001b[39m\u001b[38;5;124m'\u001b[39m, \n\u001b[0;32m     11\u001b[0m     func\u001b[38;5;241m=\u001b[39m\u001b[38;5;28;01mlambda\u001b[39;00m x: x[subset_norm_choice_cols]\u001b[38;5;241m.\u001b[39mvalue_counts()\u001b[38;5;241m.\u001b[39mmax())\n\u001b[0;32m     12\u001b[0m answer_values[\u001b[38;5;124m'\u001b[39m\u001b[38;5;124mcorrect_letter\u001b[39m\u001b[38;5;124m'\u001b[39m] \u001b[38;5;241m=\u001b[39m moe_llm_answers_comb_df[\u001b[38;5;124m'\u001b[39m\u001b[38;5;124mbench\u001b[39m\u001b[38;5;124m'\u001b[39m][\u001b[38;5;124m'\u001b[39m\u001b[38;5;124mnormalized_correct_letter\u001b[39m\u001b[38;5;124m'\u001b[39m]\n",
      "File \u001b[1;32mc:\\Users\\JamesHuckle\\anaconda3\\Lib\\site-packages\\pandas\\core\\frame.py:10034\u001b[0m, in \u001b[0;36mDataFrame.apply\u001b[1;34m(self, func, axis, raw, result_type, args, by_row, **kwargs)\u001b[0m\n\u001b[0;32m  10022\u001b[0m \u001b[38;5;28;01mfrom\u001b[39;00m \u001b[38;5;21;01mpandas\u001b[39;00m\u001b[38;5;21;01m.\u001b[39;00m\u001b[38;5;21;01mcore\u001b[39;00m\u001b[38;5;21;01m.\u001b[39;00m\u001b[38;5;21;01mapply\u001b[39;00m \u001b[38;5;28;01mimport\u001b[39;00m frame_apply\n\u001b[0;32m  10024\u001b[0m op \u001b[38;5;241m=\u001b[39m frame_apply(\n\u001b[0;32m  10025\u001b[0m     \u001b[38;5;28mself\u001b[39m,\n\u001b[0;32m  10026\u001b[0m     func\u001b[38;5;241m=\u001b[39mfunc,\n\u001b[1;32m   (...)\u001b[0m\n\u001b[0;32m  10032\u001b[0m     kwargs\u001b[38;5;241m=\u001b[39mkwargs,\n\u001b[0;32m  10033\u001b[0m )\n\u001b[1;32m> 10034\u001b[0m \u001b[38;5;28;01mreturn\u001b[39;00m \u001b[43mop\u001b[49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mapply\u001b[49m\u001b[43m(\u001b[49m\u001b[43m)\u001b[49m\u001b[38;5;241m.\u001b[39m__finalize__(\u001b[38;5;28mself\u001b[39m, method\u001b[38;5;241m=\u001b[39m\u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mapply\u001b[39m\u001b[38;5;124m\"\u001b[39m)\n",
      "File \u001b[1;32mc:\\Users\\JamesHuckle\\anaconda3\\Lib\\site-packages\\pandas\\core\\apply.py:837\u001b[0m, in \u001b[0;36mFrameApply.apply\u001b[1;34m(self)\u001b[0m\n\u001b[0;32m    834\u001b[0m \u001b[38;5;28;01melif\u001b[39;00m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39mraw:\n\u001b[0;32m    835\u001b[0m     \u001b[38;5;28;01mreturn\u001b[39;00m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39mapply_raw()\n\u001b[1;32m--> 837\u001b[0m \u001b[38;5;28;01mreturn\u001b[39;00m \u001b[38;5;28;43mself\u001b[39;49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mapply_standard\u001b[49m\u001b[43m(\u001b[49m\u001b[43m)\u001b[49m\n",
      "File \u001b[1;32mc:\\Users\\JamesHuckle\\anaconda3\\Lib\\site-packages\\pandas\\core\\apply.py:965\u001b[0m, in \u001b[0;36mFrameApply.apply_standard\u001b[1;34m(self)\u001b[0m\n\u001b[0;32m    964\u001b[0m \u001b[38;5;28;01mdef\u001b[39;00m \u001b[38;5;21mapply_standard\u001b[39m(\u001b[38;5;28mself\u001b[39m):\n\u001b[1;32m--> 965\u001b[0m     results, res_index \u001b[38;5;241m=\u001b[39m \u001b[38;5;28;43mself\u001b[39;49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mapply_series_generator\u001b[49m\u001b[43m(\u001b[49m\u001b[43m)\u001b[49m\n\u001b[0;32m    967\u001b[0m     \u001b[38;5;66;03m# wrap results\u001b[39;00m\n\u001b[0;32m    968\u001b[0m     \u001b[38;5;28;01mreturn\u001b[39;00m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39mwrap_results(results, res_index)\n",
      "File \u001b[1;32mc:\\Users\\JamesHuckle\\anaconda3\\Lib\\site-packages\\pandas\\core\\apply.py:981\u001b[0m, in \u001b[0;36mFrameApply.apply_series_generator\u001b[1;34m(self)\u001b[0m\n\u001b[0;32m    978\u001b[0m \u001b[38;5;28;01mwith\u001b[39;00m option_context(\u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mmode.chained_assignment\u001b[39m\u001b[38;5;124m\"\u001b[39m, \u001b[38;5;28;01mNone\u001b[39;00m):\n\u001b[0;32m    979\u001b[0m     \u001b[38;5;28;01mfor\u001b[39;00m i, v \u001b[38;5;129;01min\u001b[39;00m \u001b[38;5;28menumerate\u001b[39m(series_gen):\n\u001b[0;32m    980\u001b[0m         \u001b[38;5;66;03m# ignore SettingWithCopy here in case the user mutates\u001b[39;00m\n\u001b[1;32m--> 981\u001b[0m         results[i] \u001b[38;5;241m=\u001b[39m \u001b[38;5;28;43mself\u001b[39;49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mfunc\u001b[49m\u001b[43m(\u001b[49m\u001b[43mv\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[38;5;241;43m*\u001b[39;49m\u001b[38;5;28;43mself\u001b[39;49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43margs\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[38;5;241;43m*\u001b[39;49m\u001b[38;5;241;43m*\u001b[39;49m\u001b[38;5;28;43mself\u001b[39;49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mkwargs\u001b[49m\u001b[43m)\u001b[49m\n\u001b[0;32m    982\u001b[0m         \u001b[38;5;28;01mif\u001b[39;00m \u001b[38;5;28misinstance\u001b[39m(results[i], ABCSeries):\n\u001b[0;32m    983\u001b[0m             \u001b[38;5;66;03m# If we have a view on v, we need to make a copy because\u001b[39;00m\n\u001b[0;32m    984\u001b[0m             \u001b[38;5;66;03m#  series_generator will swap out the underlying data\u001b[39;00m\n\u001b[0;32m    985\u001b[0m             results[i] \u001b[38;5;241m=\u001b[39m results[i]\u001b[38;5;241m.\u001b[39mcopy(deep\u001b[38;5;241m=\u001b[39m\u001b[38;5;28;01mFalse\u001b[39;00m)\n",
      "Cell \u001b[1;32mIn[15], line 9\u001b[0m, in \u001b[0;36m<lambda>\u001b[1;34m(x)\u001b[0m\n\u001b[0;32m      5\u001b[0m subset_norm_choice_cols \u001b[38;5;241m=\u001b[39m [col \u001b[38;5;28;01mfor\u001b[39;00m col \u001b[38;5;129;01min\u001b[39;00m norm_choice_cols \u001b[38;5;28;01mif\u001b[39;00m col[\u001b[38;5;241m0\u001b[39m] \u001b[38;5;129;01min\u001b[39;00m models]\n\u001b[0;32m      6\u001b[0m answer_values \u001b[38;5;241m=\u001b[39m moe_llm_answers_comb_df\u001b[38;5;241m.\u001b[39mapply(axis\u001b[38;5;241m=\u001b[39m\u001b[38;5;124m'\u001b[39m\u001b[38;5;124mcolumns\u001b[39m\u001b[38;5;124m'\u001b[39m, \n\u001b[0;32m      7\u001b[0m     func\u001b[38;5;241m=\u001b[39m\u001b[38;5;28;01mlambda\u001b[39;00m x: x[subset_norm_choice_cols]\u001b[38;5;241m.\u001b[39mvalue_counts())\n\u001b[0;32m      8\u001b[0m most_common_answer \u001b[38;5;241m=\u001b[39m moe_llm_answers_comb_df\u001b[38;5;241m.\u001b[39mapply(axis\u001b[38;5;241m=\u001b[39m\u001b[38;5;124m'\u001b[39m\u001b[38;5;124mcolumns\u001b[39m\u001b[38;5;124m'\u001b[39m, \n\u001b[1;32m----> 9\u001b[0m     func\u001b[38;5;241m=\u001b[39m\u001b[38;5;28;01mlambda\u001b[39;00m x: get_most_common_or_random(\u001b[43mx\u001b[49m\u001b[43m[\u001b[49m\u001b[43msubset_norm_choice_cols\u001b[49m\u001b[43m]\u001b[49m), result_type\u001b[38;5;241m=\u001b[39m\u001b[38;5;124m'\u001b[39m\u001b[38;5;124mexpand\u001b[39m\u001b[38;5;124m'\u001b[39m)\n\u001b[0;32m     10\u001b[0m largest_common_answer \u001b[38;5;241m=\u001b[39m moe_llm_answers_comb_df\u001b[38;5;241m.\u001b[39mapply(axis\u001b[38;5;241m=\u001b[39m\u001b[38;5;124m'\u001b[39m\u001b[38;5;124mcolumns\u001b[39m\u001b[38;5;124m'\u001b[39m, \n\u001b[0;32m     11\u001b[0m     func\u001b[38;5;241m=\u001b[39m\u001b[38;5;28;01mlambda\u001b[39;00m x: x[subset_norm_choice_cols]\u001b[38;5;241m.\u001b[39mvalue_counts()\u001b[38;5;241m.\u001b[39mmax())\n\u001b[0;32m     12\u001b[0m answer_values[\u001b[38;5;124m'\u001b[39m\u001b[38;5;124mcorrect_letter\u001b[39m\u001b[38;5;124m'\u001b[39m] \u001b[38;5;241m=\u001b[39m moe_llm_answers_comb_df[\u001b[38;5;124m'\u001b[39m\u001b[38;5;124mbench\u001b[39m\u001b[38;5;124m'\u001b[39m][\u001b[38;5;124m'\u001b[39m\u001b[38;5;124mnormalized_correct_letter\u001b[39m\u001b[38;5;124m'\u001b[39m]\n",
      "File \u001b[1;32mc:\\Users\\JamesHuckle\\anaconda3\\Lib\\site-packages\\pandas\\core\\series.py:1072\u001b[0m, in \u001b[0;36mSeries.__getitem__\u001b[1;34m(self, key)\u001b[0m\n\u001b[0;32m   1069\u001b[0m     key \u001b[38;5;241m=\u001b[39m np\u001b[38;5;241m.\u001b[39masarray(key, dtype\u001b[38;5;241m=\u001b[39m\u001b[38;5;28mbool\u001b[39m)\n\u001b[0;32m   1070\u001b[0m     \u001b[38;5;28;01mreturn\u001b[39;00m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39m_get_rows_with_mask(key)\n\u001b[1;32m-> 1072\u001b[0m \u001b[38;5;28;01mreturn\u001b[39;00m \u001b[38;5;28;43mself\u001b[39;49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43m_get_with\u001b[49m\u001b[43m(\u001b[49m\u001b[43mkey\u001b[49m\u001b[43m)\u001b[49m\n",
      "File \u001b[1;32mc:\\Users\\JamesHuckle\\anaconda3\\Lib\\site-packages\\pandas\\core\\series.py:1113\u001b[0m, in \u001b[0;36mSeries._get_with\u001b[1;34m(self, key)\u001b[0m\n\u001b[0;32m   1110\u001b[0m         \u001b[38;5;28;01mreturn\u001b[39;00m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39miloc[key]\n\u001b[0;32m   1112\u001b[0m \u001b[38;5;66;03m# handle the dup indexing case GH#4246\u001b[39;00m\n\u001b[1;32m-> 1113\u001b[0m \u001b[38;5;28;01mreturn\u001b[39;00m \u001b[38;5;28;43mself\u001b[39;49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mloc\u001b[49m\u001b[43m[\u001b[49m\u001b[43mkey\u001b[49m\u001b[43m]\u001b[49m\n",
      "File \u001b[1;32mc:\\Users\\JamesHuckle\\anaconda3\\Lib\\site-packages\\pandas\\core\\indexing.py:1153\u001b[0m, in \u001b[0;36m_LocationIndexer.__getitem__\u001b[1;34m(self, key)\u001b[0m\n\u001b[0;32m   1150\u001b[0m axis \u001b[38;5;241m=\u001b[39m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39maxis \u001b[38;5;129;01mor\u001b[39;00m \u001b[38;5;241m0\u001b[39m\n\u001b[0;32m   1152\u001b[0m maybe_callable \u001b[38;5;241m=\u001b[39m com\u001b[38;5;241m.\u001b[39mapply_if_callable(key, \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39mobj)\n\u001b[1;32m-> 1153\u001b[0m \u001b[38;5;28;01mreturn\u001b[39;00m \u001b[38;5;28;43mself\u001b[39;49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43m_getitem_axis\u001b[49m\u001b[43m(\u001b[49m\u001b[43mmaybe_callable\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43maxis\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[43maxis\u001b[49m\u001b[43m)\u001b[49m\n",
      "File \u001b[1;32mc:\\Users\\JamesHuckle\\anaconda3\\Lib\\site-packages\\pandas\\core\\indexing.py:1382\u001b[0m, in \u001b[0;36m_LocIndexer._getitem_axis\u001b[1;34m(self, key, axis)\u001b[0m\n\u001b[0;32m   1379\u001b[0m     \u001b[38;5;28;01mif\u001b[39;00m \u001b[38;5;28mhasattr\u001b[39m(key, \u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mndim\u001b[39m\u001b[38;5;124m\"\u001b[39m) \u001b[38;5;129;01mand\u001b[39;00m key\u001b[38;5;241m.\u001b[39mndim \u001b[38;5;241m>\u001b[39m \u001b[38;5;241m1\u001b[39m:\n\u001b[0;32m   1380\u001b[0m         \u001b[38;5;28;01mraise\u001b[39;00m \u001b[38;5;167;01mValueError\u001b[39;00m(\u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mCannot index with multidimensional key\u001b[39m\u001b[38;5;124m\"\u001b[39m)\n\u001b[1;32m-> 1382\u001b[0m     \u001b[38;5;28;01mreturn\u001b[39;00m \u001b[38;5;28;43mself\u001b[39;49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43m_getitem_iterable\u001b[49m\u001b[43m(\u001b[49m\u001b[43mkey\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43maxis\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[43maxis\u001b[49m\u001b[43m)\u001b[49m\n\u001b[0;32m   1384\u001b[0m \u001b[38;5;66;03m# nested tuple slicing\u001b[39;00m\n\u001b[0;32m   1385\u001b[0m \u001b[38;5;28;01mif\u001b[39;00m is_nested_tuple(key, labels):\n",
      "File \u001b[1;32mc:\\Users\\JamesHuckle\\anaconda3\\Lib\\site-packages\\pandas\\core\\indexing.py:1322\u001b[0m, in \u001b[0;36m_LocIndexer._getitem_iterable\u001b[1;34m(self, key, axis)\u001b[0m\n\u001b[0;32m   1319\u001b[0m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39m_validate_key(key, axis)\n\u001b[0;32m   1321\u001b[0m \u001b[38;5;66;03m# A collection of keys\u001b[39;00m\n\u001b[1;32m-> 1322\u001b[0m keyarr, indexer \u001b[38;5;241m=\u001b[39m \u001b[38;5;28;43mself\u001b[39;49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43m_get_listlike_indexer\u001b[49m\u001b[43m(\u001b[49m\u001b[43mkey\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43maxis\u001b[49m\u001b[43m)\u001b[49m\n\u001b[0;32m   1323\u001b[0m \u001b[38;5;28;01mreturn\u001b[39;00m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39mobj\u001b[38;5;241m.\u001b[39m_reindex_with_indexers(\n\u001b[0;32m   1324\u001b[0m     {axis: [keyarr, indexer]}, copy\u001b[38;5;241m=\u001b[39m\u001b[38;5;28;01mTrue\u001b[39;00m, allow_dups\u001b[38;5;241m=\u001b[39m\u001b[38;5;28;01mTrue\u001b[39;00m\n\u001b[0;32m   1325\u001b[0m )\n",
      "File \u001b[1;32mc:\\Users\\JamesHuckle\\anaconda3\\Lib\\site-packages\\pandas\\core\\indexing.py:1520\u001b[0m, in \u001b[0;36m_LocIndexer._get_listlike_indexer\u001b[1;34m(self, key, axis)\u001b[0m\n\u001b[0;32m   1517\u001b[0m ax \u001b[38;5;241m=\u001b[39m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39mobj\u001b[38;5;241m.\u001b[39m_get_axis(axis)\n\u001b[0;32m   1518\u001b[0m axis_name \u001b[38;5;241m=\u001b[39m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39mobj\u001b[38;5;241m.\u001b[39m_get_axis_name(axis)\n\u001b[1;32m-> 1520\u001b[0m keyarr, indexer \u001b[38;5;241m=\u001b[39m \u001b[43max\u001b[49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43m_get_indexer_strict\u001b[49m\u001b[43m(\u001b[49m\u001b[43mkey\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43maxis_name\u001b[49m\u001b[43m)\u001b[49m\n\u001b[0;32m   1522\u001b[0m \u001b[38;5;28;01mreturn\u001b[39;00m keyarr, indexer\n",
      "File \u001b[1;32mc:\\Users\\JamesHuckle\\anaconda3\\Lib\\site-packages\\pandas\\core\\indexes\\multi.py:2651\u001b[0m, in \u001b[0;36mMultiIndex._get_indexer_strict\u001b[1;34m(self, key, axis_name)\u001b[0m\n\u001b[0;32m   2648\u001b[0m     \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39m_raise_if_missing(key, indexer, axis_name)\n\u001b[0;32m   2649\u001b[0m     \u001b[38;5;28;01mreturn\u001b[39;00m \u001b[38;5;28mself\u001b[39m[indexer], indexer\n\u001b[1;32m-> 2651\u001b[0m \u001b[38;5;28;01mreturn\u001b[39;00m \u001b[38;5;28;43msuper\u001b[39;49m\u001b[43m(\u001b[49m\u001b[43m)\u001b[49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43m_get_indexer_strict\u001b[49m\u001b[43m(\u001b[49m\u001b[43mkey\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43maxis_name\u001b[49m\u001b[43m)\u001b[49m\n",
      "File \u001b[1;32mc:\\Users\\JamesHuckle\\anaconda3\\Lib\\site-packages\\pandas\\core\\indexes\\base.py:6110\u001b[0m, in \u001b[0;36mIndex._get_indexer_strict\u001b[1;34m(self, key, axis_name)\u001b[0m\n\u001b[0;32m   6107\u001b[0m     keyarr \u001b[38;5;241m=\u001b[39m com\u001b[38;5;241m.\u001b[39masarray_tuplesafe(keyarr)\n\u001b[0;32m   6109\u001b[0m \u001b[38;5;28;01mif\u001b[39;00m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39m_index_as_unique:\n\u001b[1;32m-> 6110\u001b[0m     indexer \u001b[38;5;241m=\u001b[39m \u001b[38;5;28;43mself\u001b[39;49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mget_indexer_for\u001b[49m\u001b[43m(\u001b[49m\u001b[43mkeyarr\u001b[49m\u001b[43m)\u001b[49m\n\u001b[0;32m   6111\u001b[0m     keyarr \u001b[38;5;241m=\u001b[39m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39mreindex(keyarr)[\u001b[38;5;241m0\u001b[39m]\n\u001b[0;32m   6112\u001b[0m \u001b[38;5;28;01melse\u001b[39;00m:\n",
      "File \u001b[1;32mc:\\Users\\JamesHuckle\\anaconda3\\Lib\\site-packages\\pandas\\core\\indexes\\base.py:6097\u001b[0m, in \u001b[0;36mIndex.get_indexer_for\u001b[1;34m(self, target)\u001b[0m\n\u001b[0;32m   6079\u001b[0m \u001b[38;5;250m\u001b[39m\u001b[38;5;124;03m\"\"\"\u001b[39;00m\n\u001b[0;32m   6080\u001b[0m \u001b[38;5;124;03mGuaranteed return of an indexer even when non-unique.\u001b[39;00m\n\u001b[0;32m   6081\u001b[0m \n\u001b[1;32m   (...)\u001b[0m\n\u001b[0;32m   6094\u001b[0m \u001b[38;5;124;03marray([0, 2])\u001b[39;00m\n\u001b[0;32m   6095\u001b[0m \u001b[38;5;124;03m\"\"\"\u001b[39;00m\n\u001b[0;32m   6096\u001b[0m \u001b[38;5;28;01mif\u001b[39;00m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39m_index_as_unique:\n\u001b[1;32m-> 6097\u001b[0m     \u001b[38;5;28;01mreturn\u001b[39;00m \u001b[38;5;28;43mself\u001b[39;49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mget_indexer\u001b[49m\u001b[43m(\u001b[49m\u001b[43mtarget\u001b[49m\u001b[43m)\u001b[49m\n\u001b[0;32m   6098\u001b[0m indexer, _ \u001b[38;5;241m=\u001b[39m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39mget_indexer_non_unique(target)\n\u001b[0;32m   6099\u001b[0m \u001b[38;5;28;01mreturn\u001b[39;00m indexer\n",
      "File \u001b[1;32mc:\\Users\\JamesHuckle\\anaconda3\\Lib\\site-packages\\pandas\\core\\indexes\\base.py:3924\u001b[0m, in \u001b[0;36mIndex.get_indexer\u001b[1;34m(self, target, method, limit, tolerance)\u001b[0m\n\u001b[0;32m   3922\u001b[0m pself, ptarget \u001b[38;5;241m=\u001b[39m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39m_maybe_promote(target)\n\u001b[0;32m   3923\u001b[0m \u001b[38;5;28;01mif\u001b[39;00m pself \u001b[38;5;129;01mis\u001b[39;00m \u001b[38;5;129;01mnot\u001b[39;00m \u001b[38;5;28mself\u001b[39m \u001b[38;5;129;01mor\u001b[39;00m ptarget \u001b[38;5;129;01mis\u001b[39;00m \u001b[38;5;129;01mnot\u001b[39;00m target:\n\u001b[1;32m-> 3924\u001b[0m     \u001b[38;5;28;01mreturn\u001b[39;00m \u001b[43mpself\u001b[49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mget_indexer\u001b[49m\u001b[43m(\u001b[49m\n\u001b[0;32m   3925\u001b[0m \u001b[43m        \u001b[49m\u001b[43mptarget\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43mmethod\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[43mmethod\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43mlimit\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[43mlimit\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43mtolerance\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[43mtolerance\u001b[49m\n\u001b[0;32m   3926\u001b[0m \u001b[43m    \u001b[49m\u001b[43m)\u001b[49m\n\u001b[0;32m   3928\u001b[0m \u001b[38;5;28;01mif\u001b[39;00m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39mdtype \u001b[38;5;241m==\u001b[39m target\u001b[38;5;241m.\u001b[39mdtype \u001b[38;5;129;01mand\u001b[39;00m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39mequals(target):\n\u001b[0;32m   3929\u001b[0m     \u001b[38;5;66;03m# Only call equals if we have same dtype to avoid inference/casting\u001b[39;00m\n\u001b[0;32m   3930\u001b[0m     \u001b[38;5;28;01mreturn\u001b[39;00m np\u001b[38;5;241m.\u001b[39marange(\u001b[38;5;28mlen\u001b[39m(target), dtype\u001b[38;5;241m=\u001b[39mnp\u001b[38;5;241m.\u001b[39mintp)\n",
      "File \u001b[1;32mc:\\Users\\JamesHuckle\\anaconda3\\Lib\\site-packages\\pandas\\core\\indexes\\base.py:3943\u001b[0m, in \u001b[0;36mIndex.get_indexer\u001b[1;34m(self, target, method, limit, tolerance)\u001b[0m\n\u001b[0;32m   3938\u001b[0m     target \u001b[38;5;241m=\u001b[39m target\u001b[38;5;241m.\u001b[39mastype(dtype, copy\u001b[38;5;241m=\u001b[39m\u001b[38;5;28;01mFalse\u001b[39;00m)\n\u001b[0;32m   3939\u001b[0m     \u001b[38;5;28;01mreturn\u001b[39;00m this\u001b[38;5;241m.\u001b[39m_get_indexer(\n\u001b[0;32m   3940\u001b[0m         target, method\u001b[38;5;241m=\u001b[39mmethod, limit\u001b[38;5;241m=\u001b[39mlimit, tolerance\u001b[38;5;241m=\u001b[39mtolerance\n\u001b[0;32m   3941\u001b[0m     )\n\u001b[1;32m-> 3943\u001b[0m \u001b[38;5;28;01mreturn\u001b[39;00m \u001b[38;5;28;43mself\u001b[39;49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43m_get_indexer\u001b[49m\u001b[43m(\u001b[49m\u001b[43mtarget\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43mmethod\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43mlimit\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43mtolerance\u001b[49m\u001b[43m)\u001b[49m\n",
      "File \u001b[1;32mc:\\Users\\JamesHuckle\\anaconda3\\Lib\\site-packages\\pandas\\core\\indexes\\base.py:3964\u001b[0m, in \u001b[0;36mIndex._get_indexer\u001b[1;34m(self, target, method, limit, tolerance)\u001b[0m\n\u001b[0;32m   3961\u001b[0m     engine \u001b[38;5;241m=\u001b[39m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39m_engine\n\u001b[0;32m   3962\u001b[0m     \u001b[38;5;66;03m# error: Item \"IndexEngine\" of \"Union[IndexEngine, ExtensionEngine]\"\u001b[39;00m\n\u001b[0;32m   3963\u001b[0m     \u001b[38;5;66;03m# has no attribute \"_extract_level_codes\"\u001b[39;00m\n\u001b[1;32m-> 3964\u001b[0m     tgt_values \u001b[38;5;241m=\u001b[39m \u001b[43mengine\u001b[49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43m_extract_level_codes\u001b[49m\u001b[43m(\u001b[49m\u001b[43m  \u001b[49m\u001b[38;5;66;43;03m# type: ignore[union-attr]\u001b[39;49;00m\n\u001b[0;32m   3965\u001b[0m \u001b[43m        \u001b[49m\u001b[43mtarget\u001b[49m\n\u001b[0;32m   3966\u001b[0m \u001b[43m    \u001b[49m\u001b[43m)\u001b[49m\n\u001b[0;32m   3967\u001b[0m \u001b[38;5;28;01melse\u001b[39;00m:\n\u001b[0;32m   3968\u001b[0m     tgt_values \u001b[38;5;241m=\u001b[39m target\u001b[38;5;241m.\u001b[39m_get_engine_target()\n",
      "File \u001b[1;32mindex.pyx:732\u001b[0m, in \u001b[0;36mpandas._libs.index.BaseMultiIndexCodesEngine._extract_level_codes\u001b[1;34m()\u001b[0m\n",
      "File \u001b[1;32mc:\\Users\\JamesHuckle\\anaconda3\\Lib\\site-packages\\pandas\\core\\indexes\\multi.py:134\u001b[0m, in \u001b[0;36mMultiIndexUIntEngine._codes_to_ints\u001b[1;34m(self, codes)\u001b[0m\n\u001b[0;32m    127\u001b[0m \u001b[38;5;250m\u001b[39m\u001b[38;5;124;03m\"\"\"\u001b[39;00m\n\u001b[0;32m    128\u001b[0m \u001b[38;5;124;03mThis class manages a MultiIndex by mapping label combinations to positive\u001b[39;00m\n\u001b[0;32m    129\u001b[0m \u001b[38;5;124;03mintegers.\u001b[39;00m\n\u001b[0;32m    130\u001b[0m \u001b[38;5;124;03m\"\"\"\u001b[39;00m\n\u001b[0;32m    132\u001b[0m _base \u001b[38;5;241m=\u001b[39m libindex\u001b[38;5;241m.\u001b[39mUInt64Engine\n\u001b[1;32m--> 134\u001b[0m \u001b[38;5;28;01mdef\u001b[39;00m \u001b[38;5;21m_codes_to_ints\u001b[39m(\u001b[38;5;28mself\u001b[39m, codes):\n\u001b[0;32m    135\u001b[0m \u001b[38;5;250m    \u001b[39m\u001b[38;5;124;03m\"\"\"\u001b[39;00m\n\u001b[0;32m    136\u001b[0m \u001b[38;5;124;03m    Transform combination(s) of uint64 in one uint64 (each), in a strictly\u001b[39;00m\n\u001b[0;32m    137\u001b[0m \u001b[38;5;124;03m    monotonic way (i.e. respecting the lexicographic order of integer\u001b[39;00m\n\u001b[1;32m   (...)\u001b[0m\n\u001b[0;32m    148\u001b[0m \u001b[38;5;124;03m        Integer(s) representing one combination (each).\u001b[39;00m\n\u001b[0;32m    149\u001b[0m \u001b[38;5;124;03m    \"\"\"\u001b[39;00m\n\u001b[0;32m    150\u001b[0m     \u001b[38;5;66;03m# Shift the representation of each level by the pre-calculated number\u001b[39;00m\n\u001b[0;32m    151\u001b[0m     \u001b[38;5;66;03m# of bits:\u001b[39;00m\n",
      "\u001b[1;31mKeyboardInterrupt\u001b[0m: "
     ]
    }
   ],
   "source": [
    "from tqdm import tqdm\n",
    "\n",
    "all_combo_results = {}\n",
    "for models in tqdm(all_combos):\n",
    "    subset_norm_choice_cols = [col for col in norm_choice_cols if col[0] in models]\n",
    "    answer_values = moe_llm_answers_comb_df.apply(axis='columns', \n",
    "        func=lambda x: x[subset_norm_choice_cols].value_counts())\n",
    "    most_common_answer = moe_llm_answers_comb_df.apply(axis='columns', \n",
    "        func=lambda x: get_most_common_or_random(x[subset_norm_choice_cols]), result_type='expand')\n",
    "    largest_common_answer = moe_llm_answers_comb_df.apply(axis='columns', \n",
    "        func=lambda x: x[subset_norm_choice_cols].value_counts().max())\n",
    "    answer_values['correct_letter'] = moe_llm_answers_comb_df['bench']['normalized_correct_letter']\n",
    "    answer_values['most_common_answer'] = most_common_answer\n",
    "    answer_values['largest_common_answer'] = largest_common_answer\n",
    "    #display(answer_values)\n",
    "\n",
    "    correct = (answer_values['correct_letter'] == answer_values['most_common_answer']).sum() / len(answer_values)\n",
    "    #print(f'Models: {models} | Score %: {correct}')\n",
    "    all_combo_results[models] = correct"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Models: ('claude-3-5-sonnet-20240620', 'claude-3-opus-20240229') | Score %: 0.4666666666666667\n",
      "Models: ('claude-3-5-sonnet-20240620', 'gpt-4o') | Score %: 0.47\n",
      "Models: ('claude-3-5-sonnet-20240620', 'claude-3-opus-20240229', 'gemini-1_5-pro') | Score %: 0.4766666666666667\n",
      "Models: ('claude-3-5-sonnet-20240620', 'gemini-1_5-pro', 'gpt-4o') | Score %: 0.4866666666666667\n",
      "Models: ('claude-3-5-sonnet-20240620', 'claude-3-opus-20240229', 'gemini-1_5-pro', 'gpt-4o', 'open-mixtral-8x22b') | Score %: 0.49\n"
     ]
    }
   ],
   "source": [
    "max_score = 0\n",
    "for (models, score) in all_combo_results.items():\n",
    "    if score > max_score:\n",
    "        max_score = score\n",
    "        print(f'Models: {models} | Score %: {score}')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "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></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th>score</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th rowspan=\"5\" valign=\"top\">claude-3-5-sonnet-20240620</th>\n",
       "      <th>claude-3-opus-20240229</th>\n",
       "      <th>gemini-1_5-pro</th>\n",
       "      <th>gpt-4o</th>\n",
       "      <th>open-mixtral-8x22b</th>\n",
       "      <th>NaN</th>\n",
       "      <th>NaN</th>\n",
       "      <th>NaN</th>\n",
       "      <th>NaN</th>\n",
       "      <th>NaN</th>\n",
       "      <td>0.490000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>gemini-1_5-pro</th>\n",
       "      <th>gpt-4o</th>\n",
       "      <th>NaN</th>\n",
       "      <th>NaN</th>\n",
       "      <th>NaN</th>\n",
       "      <th>NaN</th>\n",
       "      <th>NaN</th>\n",
       "      <th>NaN</th>\n",
       "      <th>NaN</th>\n",
       "      <td>0.486667</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>claude-3-opus-20240229</th>\n",
       "      <th>gemini-1_5-pro</th>\n",
       "      <th>NaN</th>\n",
       "      <th>NaN</th>\n",
       "      <th>NaN</th>\n",
       "      <th>NaN</th>\n",
       "      <th>NaN</th>\n",
       "      <th>NaN</th>\n",
       "      <th>NaN</th>\n",
       "      <td>0.476667</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>gpt-4o</th>\n",
       "      <th>NaN</th>\n",
       "      <th>NaN</th>\n",
       "      <th>NaN</th>\n",
       "      <th>NaN</th>\n",
       "      <th>NaN</th>\n",
       "      <th>NaN</th>\n",
       "      <th>NaN</th>\n",
       "      <th>NaN</th>\n",
       "      <td>0.470000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>claude-3-opus-20240229</th>\n",
       "      <th>gemini-1_5-pro</th>\n",
       "      <th>gpt-4o</th>\n",
       "      <th>NaN</th>\n",
       "      <th>NaN</th>\n",
       "      <th>NaN</th>\n",
       "      <th>NaN</th>\n",
       "      <th>NaN</th>\n",
       "      <th>NaN</th>\n",
       "      <td>0.466667</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <th>...</th>\n",
       "      <th>...</th>\n",
       "      <th>...</th>\n",
       "      <th>...</th>\n",
       "      <th>...</th>\n",
       "      <th>...</th>\n",
       "      <th>...</th>\n",
       "      <th>...</th>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th rowspan=\"2\" valign=\"top\">command-r</th>\n",
       "      <th>meta_llama3-70b-instruct-v1_0</th>\n",
       "      <th>NaN</th>\n",
       "      <th>NaN</th>\n",
       "      <th>NaN</th>\n",
       "      <th>NaN</th>\n",
       "      <th>NaN</th>\n",
       "      <th>NaN</th>\n",
       "      <th>NaN</th>\n",
       "      <th>NaN</th>\n",
       "      <td>0.253333</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>gemini-1_0-pro</th>\n",
       "      <th>meta_llama3-70b-instruct-v1_0</th>\n",
       "      <th>NaN</th>\n",
       "      <th>NaN</th>\n",
       "      <th>NaN</th>\n",
       "      <th>NaN</th>\n",
       "      <th>NaN</th>\n",
       "      <th>NaN</th>\n",
       "      <th>NaN</th>\n",
       "      <td>0.246667</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th rowspan=\"2\" valign=\"top\">gemini-1_0-pro</th>\n",
       "      <th>mistral-large-latest</th>\n",
       "      <th>NaN</th>\n",
       "      <th>NaN</th>\n",
       "      <th>NaN</th>\n",
       "      <th>NaN</th>\n",
       "      <th>NaN</th>\n",
       "      <th>NaN</th>\n",
       "      <th>NaN</th>\n",
       "      <th>NaN</th>\n",
       "      <td>0.246667</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>meta_llama3-70b-instruct-v1_0</th>\n",
       "      <th>NaN</th>\n",
       "      <th>NaN</th>\n",
       "      <th>NaN</th>\n",
       "      <th>NaN</th>\n",
       "      <th>NaN</th>\n",
       "      <th>NaN</th>\n",
       "      <th>NaN</th>\n",
       "      <th>NaN</th>\n",
       "      <td>0.236667</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>command-r</th>\n",
       "      <th>gemini-1_0-pro</th>\n",
       "      <th>NaN</th>\n",
       "      <th>NaN</th>\n",
       "      <th>NaN</th>\n",
       "      <th>NaN</th>\n",
       "      <th>NaN</th>\n",
       "      <th>NaN</th>\n",
       "      <th>NaN</th>\n",
       "      <th>NaN</th>\n",
       "      <td>0.236667</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>1013 rows × 1 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                                                                                                                                         score\n",
       "claude-3-5-sonnet-20240620 claude-3-opus-20240229        gemini-1_5-pro                gpt-4o open-mixtral-8x22b NaN NaN NaN NaN NaN  0.490000\n",
       "                           gemini-1_5-pro                gpt-4o                        NaN    NaN                NaN NaN NaN NaN NaN  0.486667\n",
       "                           claude-3-opus-20240229        gemini-1_5-pro                NaN    NaN                NaN NaN NaN NaN NaN  0.476667\n",
       "                           gpt-4o                        NaN                           NaN    NaN                NaN NaN NaN NaN NaN  0.470000\n",
       "                           claude-3-opus-20240229        gemini-1_5-pro                gpt-4o NaN                NaN NaN NaN NaN NaN  0.466667\n",
       "...                                                                                                                                        ...\n",
       "command-r                  meta_llama3-70b-instruct-v1_0 NaN                           NaN    NaN                NaN NaN NaN NaN NaN  0.253333\n",
       "                           gemini-1_0-pro                meta_llama3-70b-instruct-v1_0 NaN    NaN                NaN NaN NaN NaN NaN  0.246667\n",
       "gemini-1_0-pro             mistral-large-latest          NaN                           NaN    NaN                NaN NaN NaN NaN NaN  0.246667\n",
       "                           meta_llama3-70b-instruct-v1_0 NaN                           NaN    NaN                NaN NaN NaN NaN NaN  0.236667\n",
       "command-r                  gemini-1_0-pro                NaN                           NaN    NaN                NaN NaN NaN NaN NaN  0.236667\n",
       "\n",
       "[1013 rows x 1 columns]"
      ]
     },
     "execution_count": 136,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pd.DataFrame(all_combo_results, index=['score']).T.sort_values(by='score', ascending=False)#.to_csv('./all_combo_results.csv')"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "base",
   "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": "3.11.7"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}