from typing import Any, List from evalscope.benchmarks import Benchmark, DataAdapter from evalscope.metrics import LLMJudge, Metric, mean, metric_registry from evalscope.utils.logger import get_logger # flake8: noqa logger = get_logger() GRADER_SYSTEM_PROMPT = "Please act as an impartial judge and evaluate the quality of the responses provided by two AI assistants to the user prompt displayed below. You will be given assistant A's answer and assistant B's answer. Your job is to evaluate which assistant's answer is better.\n\nBegin your evaluation by generating your own answer to the prompt. You must provide your answers before judging any answers.\n\nWhen evaluating the assistants' answers, compare both assistants' answers with your answer. You must identify and correct any mistakes or inaccurate information.\n\nThen consider if the assistant's answers are helpful, relevant, and concise. Helpful means the answer correctly responds to the prompt or follows the instructions. Note when user prompt has any ambiguity or more than one interpretation, it is more helpful and appropriate to ask for clarifications or more information from the user than providing an answer based on assumptions. Relevant means all parts of the response closely connect or are appropriate to what is being asked. Concise means the response is clear and not verbose or excessive.\n\nThen consider the creativity and novelty of the assistant's answers when needed. Finally, identify any missing important information in the assistants' answers that would be beneficial to include when responding to the user prompt.\n\nAfter providing your explanation, you must output only one of the following choices as your final verdict with a label:\n\n1. Assistant A is significantly better: [[A>>B]]\n2. Assistant A is slightly better: [[A>B]]\n3. Tie, relatively the same: [[A=B]]\n4. Assistant B is slightly better: [[B>A]]\n5. Assistant B is significantly better: [[B>>A]]\n\nExample output: \"My final verdict is tie: [[A=B]]\"." # noqa: E501 GRADER_TEMPLATE = "<|User Prompt|>\n{question}\n\n<|The Start of Assistant A's Answer|>\n{answer_1}\n<|The End of Assistant A's Answer|>\n\n<|The Start of Assistant B's Answer|>\n{answer_2}\n<|The End of Assistant B's Answer|>".strip( ) # noqa: E501 @Benchmark.register( name='arena_hard', pretty_name='ArenaHard', tags=['Instruction-Following', 'Arena'], description= 'ArenaHard is a benchmark designed to evaluate the performance of large language models in a competitive setting, ' 'where models are pitted against each other in a series of tasks to determine their relative strengths and weaknesses. ' 'It includes a set of challenging tasks that require reasoning, understanding, and generation capabilities. ' 'Currently not support `style-controlled winrate`; the official Judge model is `gpt-4-1106-preview`, while the baseline model is `gpt-4-0314`.', # noqa: E501 dataset_id='AI-ModelScope/arena-hard-auto-v0.1', metric_list=['winrate'], few_shot_num=0, train_split=None, eval_split='test') class ArenaHardAdapter(DataAdapter): def __init__(self, *args, **kwargs): super().__init__(*args, **kwargs) # register metrics metric_registry.register(Metric(name='winrate', object=mean)) # whether to use LLM as a judge self.llm_as_a_judge = True def gen_prompt(self, input_d: dict, subset_name: str, few_shot_list: list, **kwargs) -> dict: question = input_d['question'] return self.gen_prompt_data(question) def get_gold_answer(self, input_d: dict) -> str: return input_d['prediction'] def parse_pred_result(self, result: str, raw_input_d: dict = None, **kwargs) -> str: return result.strip() def match(self, gold: str, pred: str): # simple match logger.warning(f'Please use LLMJudge to match the result for {self.name}') return None def llm_match(self, gold: Any, pred: Any, judge: LLMJudge, **kwargs) -> dict: from .utils import post_process_arenahard raw_input = kwargs.get('raw_input', None) question = raw_input['question'] # gold is baseline answer 'A', pred is model answer 'B' prompt1 = GRADER_TEMPLATE.format(question=question, answer_1=gold, answer_2=pred) # reverse the order prompt2 = GRADER_TEMPLATE.format(question=question, answer_1=pred, answer_2=gold) # get grading response game1_response = judge(prompt1, system_prompt=GRADER_SYSTEM_PROMPT) game2_response = judge(prompt2, system_prompt=GRADER_SYSTEM_PROMPT) # parse grading response res1 = post_process_arenahard(game1_response) res2 = post_process_arenahard(game2_response) return { 'model_a': 'gpt4-0314', 'model_b': 'test_model', 'games': [ { 'user_prompt': prompt1, 'judgment': game1_response, 'score': res1 }, { 'user_prompt': prompt2, 'judgment': game2_response, 'score': res2 }, ] } def compute_metric(self, review_res_list: List[dict], **kwargs) -> List[dict]: """ compute score of the model """ import pandas as pd from .utils import compute_mle_elo, get_battles_from_row, get_bootstrap_result, get_win_rate_column if isinstance(review_res_list[0], list): review_res_list = [item for sublist in review_res_list for item in sublist] battles = pd.concat([get_battles_from_row(res) for res in review_res_list]) bootstrap_online_elo = compute_mle_elo(battles) # bootstrap_elo_lu = get_bootstrap_result(battles, compute_mle_elo, 100) stats = pd.DataFrame() stats['results'] = None stats['results'] = stats['results'].astype('object') for i, model in enumerate(bootstrap_online_elo.index): # assert model in bootstrap_elo_lu.columns stats.at[i, 'model'] = model stats.at[i, 'score'] = bootstrap_online_elo[model] # stats.at[i, "lower"] = np.percentile(bootstrap_elo_lu[model], 2.5) # stats.at[i, "upper"] = np.percentile(bootstrap_elo_lu[model], 97.5) # stats['score'] = get_win_rate_column(stats, 'score', 'gpt4-0314').tolist() score = get_win_rate_column(stats, 'score', 'gpt4-0314').at['test_model'] return [{'metric_name': 'winrate', 'score': score, 'num': len(review_res_list)}]