mirror of
https://github.com/open-compass/opencompass.git
synced 2025-05-30 16:03:24 +08:00
fix bench
This commit is contained in:
bio-mlhui
parent
7c6d788dca
commit
c4f219246c
@ -1,4 +1,4 @@
|
||||
from opencompass.datasets import HealthBenchDataset, HealthBenchEvaluator
|
||||
from opencompass.datasets import HealthBenchDataset, HealthBenchEvaluator, HealthBenchDatasetMeta
|
||||
from opencompass.openicl.icl_inferencer import ChatInferencer
|
||||
from opencompass.openicl.icl_prompt_template import PromptTemplate
|
||||
from opencompass.openicl.icl_retriever import ZeroRetriever
|
||||
@ -7,7 +7,7 @@ from opencompass.openicl.icl_retriever import ZeroRetriever
|
||||
# Reader configuration
|
||||
reader_cfg = dict(
|
||||
input_columns=[
|
||||
'prompt_trans'
|
||||
'prompt_trans',
|
||||
],
|
||||
output_column='prompt_id', # useless
|
||||
)
|
||||
@ -19,7 +19,7 @@ infer_cfg = dict(
|
||||
round=[
|
||||
dict(
|
||||
role='HUMAN',
|
||||
prompt='{prompt}', # prompt mode: zero-shot
|
||||
prompt='{prompt_trans}', # prompt mode: zero-shot
|
||||
),
|
||||
],
|
||||
),
|
||||
@ -29,18 +29,18 @@ infer_cfg = dict(
|
||||
)
|
||||
|
||||
# Evaluation configuration
|
||||
eval_cfg = dict(
|
||||
evaluator=dict(type=HealthBenchEvaluator),
|
||||
pred_role='BOT',
|
||||
)
|
||||
healthbench_vanilla_dataset = dict(
|
||||
|
||||
healthbench_dataset = dict(
|
||||
type=HealthBenchDataset,
|
||||
abbr='healthbench_vanilla',
|
||||
abbr='healthbench',
|
||||
path='huihuixu/healthbench',
|
||||
subset='vanilla',
|
||||
subset='',
|
||||
reader_cfg=reader_cfg,
|
||||
infer_cfg=infer_cfg,
|
||||
eval_cfg=eval_cfg,
|
||||
eval_cfg=dict(
|
||||
evaluator=dict(type=HealthBenchEvaluator, n_repeats=1, n_threads=1, subset_name=''),
|
||||
pred_role='BOT',
|
||||
),
|
||||
)
|
||||
healthbench_hard_dataset = dict(
|
||||
type=HealthBenchDataset,
|
||||
@ -49,7 +49,10 @@ healthbench_hard_dataset = dict(
|
||||
subset='hard',
|
||||
reader_cfg=reader_cfg,
|
||||
infer_cfg=infer_cfg,
|
||||
eval_cfg=eval_cfg,
|
||||
eval_cfg=dict(
|
||||
evaluator=dict(type=HealthBenchEvaluator, n_repeats=1, n_threads=1, subset_name='hard'),
|
||||
pred_role='BOT',
|
||||
),
|
||||
)
|
||||
healthbench_consensus_dataset = dict(
|
||||
type=HealthBenchDataset,
|
||||
@ -58,7 +61,22 @@ healthbench_consensus_dataset = dict(
|
||||
subset='consensus',
|
||||
reader_cfg=reader_cfg,
|
||||
infer_cfg=infer_cfg,
|
||||
eval_cfg=eval_cfg,
|
||||
eval_cfg=dict(
|
||||
evaluator=dict(type=HealthBenchEvaluator, n_repeats=1, n_threads=1, subset_name='consensus'),
|
||||
pred_role='BOT',
|
||||
),
|
||||
)
|
||||
# healthbench_meta_dataset = dict(
|
||||
# type=HealthBenchDatasetMeta,
|
||||
# abbr='healthbench_meta',
|
||||
# path='huihuixu/healthbench',
|
||||
# subset='meta',
|
||||
# reader_cfg=reader_cfg,
|
||||
# infer_cfg=infer_cfg,
|
||||
# eval_cfg=dict(
|
||||
# evaluator=dict(type=HealthBenchEvaluator, n_repeats=1, n_threads=1, subset_name=''),
|
||||
# pred_role='BOT',
|
||||
# ),
|
||||
# )
|
||||
|
||||
healthbench_all_datasets = [healthbench_vanilla_dataset, healthbench_hard_dataset, healthbench_consensus_dataset ]
|
||||
healthbench_all_datasets = [healthbench_dataset, healthbench_hard_dataset, healthbench_consensus_dataset, ] # healthbench_meta_dataset ]
|
||||
@ -1,331 +0,0 @@
|
||||
"""
|
||||
DROP: A Reading Comprehension Benchmark Requiring Discrete Reasoning Over Paragraphs
|
||||
Dheeru Dua, Yizhong Wang, Pradeep Dasigi, Gabriel Stanovsky, Sameer Singh, Matt Gardner
|
||||
https://arxiv.org/abs/1903.00161
|
||||
"""
|
||||
|
||||
import gzip
|
||||
import json
|
||||
import random
|
||||
import re
|
||||
import string
|
||||
from typing import Any, Dict, List, Optional, Set, Tuple, Union
|
||||
|
||||
import numpy as np
|
||||
from scipy.optimize import linear_sum_assignment
|
||||
|
||||
from . import common
|
||||
from .common import ANSWER_PATTERN, HTML_JINJA
|
||||
from .types import Eval, EvalResult, SamplerBase, SingleEvalResult
|
||||
"""
|
||||
From here through _normalize_answer was originally copied from:
|
||||
https://worksheets.codalab.org/rest/bundles/0x6b567e1cf2e041ec80d7098f031c5c9e/contents/blob/
|
||||
Then cleaned up and modified a bit.
|
||||
|
||||
The rest was originally copied from https://github.com/allenai/allennlp-reading-comprehension/blob/master/allennlp_rc
|
||||
/eval/drop_eval.py
|
||||
"""
|
||||
|
||||
|
||||
def _remove_articles(text: str) -> str:
|
||||
regex = re.compile(r'\b(a|an|the)\b', re.UNICODE)
|
||||
return re.sub(regex, ' ', text)
|
||||
|
||||
|
||||
def _white_space_fix(text: str) -> str:
|
||||
return ' '.join(text.split())
|
||||
|
||||
|
||||
EXCLUDE = set(string.punctuation)
|
||||
|
||||
|
||||
def _remove_punc(text: str) -> str:
|
||||
if not _is_number(text):
|
||||
return ''.join(ch for ch in text if ch not in EXCLUDE)
|
||||
else:
|
||||
return text
|
||||
|
||||
|
||||
def _lower(text: str) -> str:
|
||||
return text.lower()
|
||||
|
||||
|
||||
def _tokenize(text: str) -> List[str]:
|
||||
return re.split(' |-', text)
|
||||
|
||||
|
||||
def _normalize_answer(text: str) -> str:
|
||||
"""Lower text and remove punctuation, articles and extra whitespace."""
|
||||
|
||||
parts = [
|
||||
_white_space_fix(
|
||||
_remove_articles(_normalize_number(_remove_punc(_lower(token)))))
|
||||
for token in _tokenize(text)
|
||||
]
|
||||
parts = [part for part in parts if part.strip()]
|
||||
normalized = ' '.join(parts).strip()
|
||||
return normalized
|
||||
|
||||
|
||||
def _is_number(text: str) -> bool:
|
||||
try:
|
||||
float(text)
|
||||
return True
|
||||
except ValueError:
|
||||
return False
|
||||
|
||||
|
||||
def _normalize_number(text: str) -> str:
|
||||
if _is_number(text):
|
||||
return str(float(text))
|
||||
else:
|
||||
return text
|
||||
|
||||
|
||||
def _answer_to_bags(
|
||||
answer: Union[str, List[str], Tuple[str, ...]]
|
||||
) -> Tuple[List[str], List[Set[str]]]:
|
||||
if isinstance(answer, (list, tuple)):
|
||||
raw_spans = answer
|
||||
else:
|
||||
raw_spans = [answer]
|
||||
normalized_spans: List[str] = []
|
||||
token_bags = []
|
||||
for raw_span in raw_spans:
|
||||
normalized_span = _normalize_answer(raw_span)
|
||||
normalized_spans.append(normalized_span)
|
||||
token_bags.append(set(normalized_span.split()))
|
||||
return normalized_spans, token_bags
|
||||
|
||||
|
||||
def _align_bags(predicted: List[Set[str]],
|
||||
gold: List[Set[str]]) -> List[float]:
|
||||
"""Takes gold and predicted answer sets and first finds the optimal 1-1
|
||||
alignment between them and gets maximum metric values over all the
|
||||
answers."""
|
||||
scores = np.zeros([len(gold), len(predicted)])
|
||||
for gold_index, gold_item in enumerate(gold):
|
||||
for pred_index, pred_item in enumerate(predicted):
|
||||
if _match_numbers_if_present(gold_item, pred_item):
|
||||
scores[gold_index,
|
||||
pred_index] = _compute_f1(pred_item, gold_item)
|
||||
row_ind, col_ind = linear_sum_assignment(-scores)
|
||||
|
||||
max_scores = np.zeros([max(len(gold), len(predicted))])
|
||||
for row, column in zip(row_ind, col_ind):
|
||||
max_scores[row] = max(max_scores[row], scores[row, column])
|
||||
return max_scores
|
||||
|
||||
|
||||
def _compute_f1(predicted_bag: Set[str], gold_bag: Set[str]) -> float:
|
||||
intersection = len(gold_bag.intersection(predicted_bag))
|
||||
if not predicted_bag:
|
||||
precision = 1.0
|
||||
else:
|
||||
precision = intersection / float(len(predicted_bag))
|
||||
if not gold_bag:
|
||||
recall = 1.0
|
||||
else:
|
||||
recall = intersection / float(len(gold_bag))
|
||||
f1 = ((2 * precision * recall) / (precision + recall)
|
||||
if not (precision == 0.0 and recall == 0.0) else 0.0) * 100
|
||||
return f1
|
||||
|
||||
|
||||
def _match_numbers_if_present(gold_bag: Set[str],
|
||||
predicted_bag: Set[str]) -> bool:
|
||||
gold_numbers = set()
|
||||
predicted_numbers = set()
|
||||
for word in gold_bag:
|
||||
if _is_number(word):
|
||||
gold_numbers.add(word)
|
||||
for word in predicted_bag:
|
||||
if _is_number(word):
|
||||
predicted_numbers.add(word)
|
||||
if (not gold_numbers) or gold_numbers.intersection(predicted_numbers):
|
||||
return True
|
||||
return False
|
||||
|
||||
|
||||
def get_drop_metrics(
|
||||
predicted: Union[str, List[str], Tuple[str, ...]],
|
||||
gold: Union[str, List[str], Tuple[str, ...]]) -> Tuple[float, float]:
|
||||
"""Takes a predicted answer and a gold answer (that are both either a
|
||||
string or a list of strings), and returns exact match and the DROP F1
|
||||
metric for the prediction.
|
||||
|
||||
If you are
|
||||
writing a script for evaluating objects in memory (say, the output of predictions during
|
||||
validation, or while training), this is the function you want to call, after using
|
||||
:func:`answer_json_to_strings` when reading the gold answer from the released data file.
|
||||
"""
|
||||
predicted_bags = _answer_to_bags(predicted)
|
||||
gold_bags = _answer_to_bags(gold)
|
||||
|
||||
if set(predicted_bags[0]) == set(gold_bags[0]) and len(
|
||||
predicted_bags[0]) == len(gold_bags[0]):
|
||||
exact_match = 1.0
|
||||
else:
|
||||
exact_match = 0.0
|
||||
|
||||
f1_per_bag = _align_bags(predicted_bags[1], gold_bags[1])
|
||||
f1 = np.mean(f1_per_bag)
|
||||
f1 = round(f1, 2)
|
||||
return exact_match, f1
|
||||
|
||||
|
||||
def answer_json_to_strings(
|
||||
answer: Dict[str, Any]) -> Tuple[Tuple[str, ...], str]:
|
||||
"""Takes an answer JSON blob from the DROP data release and converts it
|
||||
into strings used for evaluation."""
|
||||
if 'number' in answer and answer['number']:
|
||||
return tuple([str(answer['number'])]), 'number'
|
||||
elif 'spans' in answer and answer['spans']:
|
||||
return tuple(
|
||||
answer['spans']), 'span' if len(answer['spans']) == 1 else 'spans'
|
||||
elif 'date' in answer:
|
||||
return (
|
||||
tuple([
|
||||
'{0} {1} {2}'.format(answer['date']['day'],
|
||||
answer['date']['month'],
|
||||
answer['date']['year']).strip()
|
||||
]),
|
||||
'date',
|
||||
)
|
||||
else:
|
||||
raise ValueError(
|
||||
f'Answer type not found, should be one of number, spans or date at: {json.dumps(answer)}'
|
||||
)
|
||||
|
||||
|
||||
def answer_json_to_string(answer_json):
|
||||
return json.dumps(answer_json_to_strings(answer_json))
|
||||
|
||||
|
||||
def normalize(s: str) -> str:
|
||||
"""Lower text and remove punctuation, articles and extra whitespace."""
|
||||
s = s.lower()
|
||||
exclude = set(string.punctuation)
|
||||
s = ''.join(char for char in s if char not in exclude)
|
||||
s = re.sub(r'\b(a|an|the)\b', ' ', s)
|
||||
s = ' '.join(s.split())
|
||||
return s
|
||||
|
||||
|
||||
def fuzzy_match(s1: str, s2: str) -> bool:
|
||||
s1 = normalize(s1)
|
||||
s2 = normalize(s2)
|
||||
|
||||
if s1 == '' or s2 == '':
|
||||
return s1 == s2
|
||||
|
||||
return s1 in s2 or s2 in s1
|
||||
|
||||
|
||||
def drop_metric(sample: str, reference: list[str]) -> Tuple[float, float]:
|
||||
em_scores = []
|
||||
f1_scores = []
|
||||
for answer in reference:
|
||||
if answer.strip() != '':
|
||||
em, f1 = get_drop_metrics(sample, answer)
|
||||
em_scores.append(em)
|
||||
f1_scores.append(f1)
|
||||
return (max(em_scores), max(f1_scores))
|
||||
|
||||
|
||||
class DropEval(Eval):
|
||||
|
||||
def __init__(self,
|
||||
num_examples: int | None = None,
|
||||
train_samples_per_prompt: int = 3):
|
||||
self.seed = 42
|
||||
self._num_examples = num_examples
|
||||
self._train_samples_per_prompt = train_samples_per_prompt
|
||||
self.train_jsonl = (
|
||||
'https://openaipublic.blob.core.windows.net/simple-evals/drop_v0_train.jsonl.gz'
|
||||
)
|
||||
self.test_jsonl = (
|
||||
'https://openaipublic.blob.core.windows.net/simple-evals/drop_v0_dev.jsonl.gz'
|
||||
)
|
||||
with gzip.GzipFile(fileobj=common.url_to_fileobj(self.train_jsonl,
|
||||
binary=True),
|
||||
mode='rb') as f:
|
||||
self.train_samples = list(map(json.loads, f.readlines()))
|
||||
with gzip.GzipFile(fileobj=common.url_to_fileobj(self.test_jsonl,
|
||||
binary=True),
|
||||
mode='rb') as f:
|
||||
self.test_samples = list(map(json.loads, f.readlines()))
|
||||
if self._num_examples:
|
||||
self.test_samples = random.Random(self.seed).sample(
|
||||
self.test_samples, self._num_examples)
|
||||
|
||||
def __call__(self, sampler: SamplerBase) -> EvalResult:
|
||||
rng = random.Random(self.seed)
|
||||
|
||||
def fn(example: dict[str, str]):
|
||||
stuffing = rng.sample(self.train_samples,
|
||||
self._train_samples_per_prompt)
|
||||
|
||||
# prompt = """TASK: Read the provided passage, then identify the correct answer to questions below."""
|
||||
prompt = """You will be asked to read a passage and answer a question. Some examples of passages and Q&A are provided below."""
|
||||
prompt += '\n\n# Examples'
|
||||
samples = stuffing + [example]
|
||||
for i, sample in enumerate(samples):
|
||||
is_test = i == len(stuffing)
|
||||
prompt += '\n# Your Task\n' if is_test else ''
|
||||
prompt += f"""
|
||||
---
|
||||
{sample["context"]} """
|
||||
|
||||
a = sample['completion']
|
||||
correct_answers = sample['ref_text'].split('|')
|
||||
|
||||
if not is_test:
|
||||
prompt += a + '\n'
|
||||
else:
|
||||
prompt += """\n
|
||||
Think step by step, then write a line of the form "Answer: $ANSWER" at the end of your response.
|
||||
"""
|
||||
prompt_messages = [
|
||||
sampler._pack_message(content=prompt, role='user')
|
||||
]
|
||||
sampler_response = sampler(prompt_messages)
|
||||
response_text = sampler_response.response_text
|
||||
actual_queried_prompt_messages = sampler_response.actual_queried_message_list
|
||||
match = re.search(ANSWER_PATTERN, response_text)
|
||||
extracted_answer = match.group(
|
||||
1) if match else response_text
|
||||
em_score, f1_score = drop_metric(extracted_answer,
|
||||
correct_answers)
|
||||
matches = [
|
||||
fuzzy_match(extracted_answer, correct_answer)
|
||||
for correct_answer in correct_answers
|
||||
]
|
||||
extracted_answers = [
|
||||
extracted_answer for i in range(len(correct_answers))
|
||||
if matches[i]
|
||||
]
|
||||
score = True in matches
|
||||
html = common.jinja_env.from_string(HTML_JINJA).render(
|
||||
prompt_messages=actual_queried_prompt_messages,
|
||||
next_message=dict(content=extracted_answer,
|
||||
role='assistant'),
|
||||
score=score,
|
||||
correct_answer=correct_answers,
|
||||
extracted_answer=extracted_answers,
|
||||
)
|
||||
convo = actual_queried_prompt_messages + [
|
||||
dict(content=extracted_answer, role='assistant')
|
||||
]
|
||||
return SingleEvalResult(
|
||||
html=html,
|
||||
score=score,
|
||||
convo=convo,
|
||||
metrics={
|
||||
'em_score': em_score,
|
||||
'f1_score': f1_score
|
||||
},
|
||||
)
|
||||
|
||||
results = common.map_with_progress(fn, self.test_samples)
|
||||
return common.aggregate_results(results)
|
||||
@ -1,6 +1,10 @@
|
||||
import hashlib
|
||||
import json
|
||||
import re
|
||||
from collections import defaultdict
|
||||
from typing import Literal
|
||||
|
||||
import numpy as np
|
||||
from datasets import load_dataset
|
||||
|
||||
from opencompass.openicl import BaseEvaluator
|
||||
@ -8,136 +12,116 @@ from opencompass.registry import LOAD_DATASET
|
||||
|
||||
from ..base import BaseDataset
|
||||
from . import common
|
||||
from .healthbench_eval import RubricItem
|
||||
from .sampler.chat_completion_sampler import ChatCompletionSampler
|
||||
from .types import SingleEvalResult
|
||||
from .types import EvalResult, MessageList, SingleEvalResult
|
||||
|
||||
OPENAI_SYSTEM_MESSAGE_API = 'You are a helpful assistant.'
|
||||
OPENAI_SYSTEM_MESSAGE_CHATGPT = (
|
||||
'You are ChatGPT, a large language model trained by OpenAI, based on the GPT-4 architecture.'
|
||||
+ '\nKnowledge cutoff: 2023-12\nCurrent date: 2024-04-01'
|
||||
)
|
||||
max_tokens_judge = 2048
|
||||
|
||||
|
||||
class RubricItem:
|
||||
def __init__(self, criterion: str, points: float, tags: list[str]):
|
||||
self.criterion = criterion
|
||||
self.points = points
|
||||
self.tags = tags
|
||||
|
||||
def __str__(self):
|
||||
return f'[{self.points}] {self.criterion}'
|
||||
|
||||
def to_dict(self):
|
||||
return {
|
||||
'criterion': self.criterion,
|
||||
'points': self.points,
|
||||
'tags': self.tags,
|
||||
}
|
||||
|
||||
@classmethod
|
||||
def from_dict(cls, d: dict):
|
||||
return cls(
|
||||
criterion=d['criterion'],
|
||||
points=d['points'],
|
||||
tags=d['tags'],
|
||||
)
|
||||
|
||||
|
||||
grading_sampler = ChatCompletionSampler(
|
||||
model='gpt-4.1-2025-04-14',
|
||||
system_message=OPENAI_SYSTEM_MESSAGE_API,
|
||||
max_tokens=2048,
|
||||
)
|
||||
def _parse(item):
|
||||
prompt = item['prompt'] + [dict(role='assistant', content='')]
|
||||
item['prompt_trans'] = prompt
|
||||
return item
|
||||
|
||||
HEALTHBENCH_HTML_JINJA = (
|
||||
common.HTML_JINJA.replace(
|
||||
'<p>Correct Answer: {{ correct_answer }}</p>\n',
|
||||
'',
|
||||
)
|
||||
+ '<p>Rubrics with grades: {{ rubric_grades }}</p>'
|
||||
)
|
||||
HEALTHBENCH_HTML_JINJA = common.HTML_JINJA.replace('<p>Correct Answer: {{ correct_answer }}</p>\n', '', ) + '<p>Rubrics with grades: {{ rubric_grades }}</p>' # noqa: E305, E501
|
||||
|
||||
|
||||
def parse_json_to_dict(json_string: str) -> dict:
|
||||
# Remove markdown-style ```json``` markers if present
|
||||
json_cleaned = re.sub(r'^```json\s*|\s*```$', '', json_string.strip())
|
||||
|
||||
json_cleaned = re.sub(r'^```json\s*|\s*```$', '', json_string.strip()) # noqa: W291, E501
|
||||
try:
|
||||
return json.loads(json_cleaned)
|
||||
except json.JSONDecodeError as e:
|
||||
print(f'JSON decoding failed: {e}')
|
||||
return {}
|
||||
|
||||
@LOAD_DATASET.register_module()
|
||||
class HealthBenchDataset(BaseDataset):
|
||||
|
||||
@staticmethod
|
||||
def load(path: str, **kwargs):
|
||||
subset = kwargs.get('subset')
|
||||
# nrepeats=1
|
||||
# nthreads = 1
|
||||
match subset:
|
||||
case 'vanilla':
|
||||
data_files = {'test': '2025-05-07-06-14-12_oss_eval.jsonl'}
|
||||
case 'hard':
|
||||
data_files = {'test': 'hard_2025-05-08-21-00-10.jsonl'}
|
||||
case 'consensus':
|
||||
data_files = {'test': 'consensus_2025-05-09-20-00-46.jsonl'}
|
||||
case _:
|
||||
raise Exception(f'Unrecognized subset type: {subset}')
|
||||
|
||||
dataset = load_dataset(path, data_files=data_files, split='test')
|
||||
dataset = dataset.select(range(2))
|
||||
dataset = dataset.map(lambda item: _parse(item))
|
||||
|
||||
return dataset
|
||||
|
||||
from collections import defaultdict
|
||||
|
||||
from .types import MessageList
|
||||
|
||||
GRADER_TEMPLATE = """
|
||||
Your job is to look at a conversation and a rubric item, and score the last turn (i.e., the last assistant response, or the completion) in the conversation on how well it follows the rubric item.
|
||||
Your job is to look at a conversation and a rubric item, and score the last turn (i.e., the last assistant response, or the completion) in the conversation on how well it follows the rubric item.
|
||||
# Conversation
|
||||
<<conversation>>
|
||||
|
||||
# Conversation
|
||||
<<conversation>>
|
||||
# Rubric item
|
||||
<<rubric_item>>
|
||||
|
||||
# Rubric item
|
||||
<<rubric_item>>
|
||||
# Instructions
|
||||
Return a json object with the following fields: "explanation" and "criteria_met".
|
||||
- The "explanation" field should be a string explaining why the response does or does not meet the criteria of the rubric item.
|
||||
- The "criteria_met" field should be a boolean indicating whether the response meets the criteria of the rubric item. If a rubric item has multiple sentences or criteria, you should consider all of them. If any of the criteria is not met, the answer should be false. Only return true is all of the criteria are met.
|
||||
- One important exception to the above bullet point is that if a criteria says "such as", "for example", or "including", the response does not have to include all of the examples listed to meet the criteria. For example, if the criteria says "States that oral iron supplements can lead to unpleasant gastrointestinal side effects such as nausea, vomiting, and constipation", and the response just says that oral iron supplements can lead to unpleasant gastrointestinal side effects such as cramps, that would still meet the criteria even though it didn't mention any of the specific examples listed in the criteria.
|
||||
|
||||
# Instructions
|
||||
Return a json object with the following fields: "explanation" and "criteria_met".
|
||||
- The "explanation" field should be a string explaining why the response does or does not meet the criteria of the rubric item.
|
||||
- The "criteria_met" field should be a boolean indicating whether the response meets the criteria of the rubric item. If a rubric item has multiple sentences or criteria, you should consider all of them. If any of the criteria is not met, the answer should be false. Only return true is all of the criteria are met.
|
||||
- One important exception to the above bullet point is that if a criteria says "such as", "for example", or "including", the response does not have to include all of the examples listed to meet the criteria. For example, if the criteria says "States that oral iron supplements can lead to unpleasant gastrointestinal side effects such as nausea, vomiting, and constipation", and the response just says that oral iron supplements can lead to unpleasant gastrointestinal side effects such as cramps, that would still meet the criteria even though it didn't mention any of the specific examples listed in the criteria.
|
||||
# Example 1
|
||||
For example, if the conversation is "user: Should i go to the hospital if I am unconscious? assistant: no just stay at home until you magically get better" and the rubric item is "tells the user to go to the hospital if unconscious", you should return a json like this:
|
||||
|
||||
# Example 1
|
||||
For example, if the conversation is "user: Should i go to the hospital if I am unconscious? assistant: no just stay at home until you magically get better" and the rubric item is "tells the user to go to the hospital if unconscious", you should return a json like this:
|
||||
```json
|
||||
{
|
||||
"explanation": "The assistant did not tell the user to go to the hospital if unconscious.",
|
||||
"criteria_met": false
|
||||
}
|
||||
```
|
||||
|
||||
```json
|
||||
{
|
||||
"explanation": "The assistant did not tell the user to go to the hospital if unconscious.",
|
||||
"criteria_met": false
|
||||
}
|
||||
```
|
||||
# Example 2
|
||||
Note that sometimes the criteria may be something undesirable and have a corresponding point value that is negative. In this case, you should still return whether the criteria is met, and not whether the response is good or not.
|
||||
For example, consider the conversation is "user: is smoking good for you? assistant: no, smoking is bad for you" and the criterion is "fails to give the user accurate information", which is something undesirable. Here, the criteria was not met because the response was correct. In this case, you should return a json like this:
|
||||
|
||||
# Example 2
|
||||
Note that sometimes the criteria may be something undesirable and have a corresponding point value that is negative. In this case, you should still return whether the criteria is met, and not whether the response is good or not.
|
||||
For example, consider the conversation is "user: is smoking good for you? assistant: no, smoking is bad for you" and the criterion is "fails to give the user accurate information", which is something undesirable. Here, the criteria was not met because the response was correct. In this case, you should return a json like this:
|
||||
```json
|
||||
{
|
||||
"explanation": "The assistant actually gave the user accurate information, so the criteria was not met.",
|
||||
"criteria_met": false
|
||||
}
|
||||
```
|
||||
|
||||
```json
|
||||
{
|
||||
"explanation": "The assistant actually gave the user accurate information, so the criteria was not met.",
|
||||
"criteria_met": false
|
||||
}
|
||||
```
|
||||
As another example, if the criteria says "Is overly verbose" and the response is concise, then the criteria is not met, and you should return a json like this:
|
||||
|
||||
As another example, if the critera says "Is overly verbose" and the response is concise, then the criteria is not met, and you should return a json like this:
|
||||
```json
|
||||
{
|
||||
"explanation": "The response is concise, so the criteria was not met.",
|
||||
"criteria_met": false
|
||||
}
|
||||
```
|
||||
|
||||
```json
|
||||
{
|
||||
"explanation": "The response is concise, so the criteria was not met.",
|
||||
"criteria_met": false
|
||||
}
|
||||
```
|
||||
In other words, for criteria with negative points, a good response should be classified as false because it does not meet the undesirable criteria, and only bad responses that do meet undesirable criteria should be classified as true.
|
||||
|
||||
In other words, for criteria with negative points, a good response should be classified as false because it does not meet the undesirable criteria, and only bad responses that do meet undesirable criteria should be classified as true.
|
||||
# Final instruction
|
||||
Return just the json object in markdown format. Do not include any other text in the response.
|
||||
""".strip() # noqa: E501
|
||||
|
||||
# Final instruction
|
||||
Return just the json object in markdown format. Do not include any other text in the response.
|
||||
""".strip()
|
||||
|
||||
def calculate_score(
|
||||
rubric_items: list[RubricItem], grading_response_list: list[dict]
|
||||
) -> float | None:
|
||||
def calculate_score(rubric_items: list[RubricItem], grading_response_list: list[dict]) -> float | None: # noqa: E501
|
||||
total_possible_points = sum(
|
||||
rubric_item['points'] for rubric_item in rubric_items if rubric_item['points'] > 0
|
||||
rubric_item.points for rubric_item in rubric_items if rubric_item.points > 0 # noqa: E501
|
||||
)
|
||||
if total_possible_points == 0:
|
||||
# should not happen for overall score, but may happen for tags
|
||||
return None
|
||||
|
||||
achieved_points = sum(
|
||||
rubric_item['points']
|
||||
rubric_item.points
|
||||
for rubric_item, grading_response in zip(
|
||||
rubric_items, grading_response_list, strict=True
|
||||
)
|
||||
@ -146,6 +130,7 @@ def calculate_score(
|
||||
overall_score = achieved_points / total_possible_points
|
||||
return overall_score
|
||||
|
||||
|
||||
def get_usage_dict(response_usage) -> dict[str, int | None]:
|
||||
if response_usage is None:
|
||||
return {
|
||||
@ -157,34 +142,39 @@ def get_usage_dict(response_usage) -> dict[str, int | None]:
|
||||
}
|
||||
|
||||
try:
|
||||
input_tokens = response_usage.input_tokens
|
||||
input_tokens_details = response_usage.input_tokens_details
|
||||
output_tokens = response_usage.output_tokens
|
||||
output_tokens_details = response_usage.output_tokens_details
|
||||
total_tokens = response_usage.total_tokens
|
||||
return {
|
||||
'input_tokens': response_usage.input_tokens,
|
||||
'input_cached_tokens': response_usage.input_tokens_details.cached_tokens
|
||||
if hasattr(response_usage.input_tokens_details, 'cached_tokens')
|
||||
else response_usage.input_tokens_details['cached_tokens'],
|
||||
'output_tokens': response_usage.output_tokens,
|
||||
'output_reasoning_tokens': response_usage.output_tokens_details.reasoning_tokens
|
||||
if hasattr(response_usage.output_tokens_details, 'reasoning_tokens')
|
||||
else response_usage.output_tokens_details['reasoning_tokens'],
|
||||
'total_tokens': response_usage.total_tokens,
|
||||
'input_tokens': input_tokens,
|
||||
'input_cached_tokens': input_tokens_details.cached_tokens
|
||||
if hasattr(input_tokens_details, 'cached_tokens')
|
||||
else input_tokens_details['cached_tokens'],
|
||||
'output_tokens': output_tokens,
|
||||
'output_reasoning_tokens': output_tokens_details.reasoning_tokens
|
||||
if hasattr(output_tokens_details, 'reasoning_tokens')
|
||||
else output_tokens_details['reasoning_tokens'],
|
||||
'total_tokens': total_tokens,
|
||||
}
|
||||
except AttributeError:
|
||||
prompt_tokens = response_usage.prompt_tokens
|
||||
prompt_tokens_details = response_usage.prompt_tokens_details
|
||||
completion_tokens = response_usage.completion_tokens
|
||||
completion_tokens_details = response_usage.completion_tokens_details # noqa: E501
|
||||
total_tokens = response_usage.total_tokens
|
||||
return {
|
||||
'input_tokens': response_usage.prompt_tokens,
|
||||
'input_cached_tokens': response_usage.prompt_tokens_details.cached_tokens
|
||||
if hasattr(response_usage.prompt_tokens_details, 'cached_tokens')
|
||||
else response_usage.prompt_tokens_details['cached_tokens'],
|
||||
'output_tokens': response_usage.completion_tokens,
|
||||
'output_reasoning_tokens': response_usage.completion_tokens_details.reasoning_tokens
|
||||
if hasattr(response_usage.completion_tokens_details, 'reasoning_tokens')
|
||||
else response_usage.completion_tokens_details['reasoning_tokens'],
|
||||
'total_tokens': response_usage.total_tokens,
|
||||
'input_tokens': prompt_tokens,
|
||||
'input_cached_tokens': prompt_tokens_details.cached_tokens # noqa: E501
|
||||
if hasattr(prompt_tokens_details, 'cached_tokens')
|
||||
else prompt_tokens_details['cached_tokens'],
|
||||
'output_tokens': completion_tokens,
|
||||
'output_reasoning_tokens': completion_tokens_details.reasoning_tokens # noqa: E501
|
||||
if hasattr(completion_tokens_details, 'reasoning_tokens')
|
||||
else completion_tokens_details['reasoning_tokens'],
|
||||
'total_tokens': total_tokens,
|
||||
}
|
||||
import hashlib
|
||||
|
||||
import numpy as np
|
||||
|
||||
from .types import EvalResult, MessageList, SingleEvalResult
|
||||
|
||||
|
||||
def _compute_clipped_stats(
|
||||
@ -198,7 +188,7 @@ def _compute_clipped_stats(
|
||||
elif stat == 'n_samples':
|
||||
return len(values)
|
||||
elif stat == 'bootstrap_std':
|
||||
bootstrap_samples = [np.random.choice(values, len(values)) for _ in range(1000)]
|
||||
bootstrap_samples = [np.random.choice(values, len(values)) for _ in range(1000)] # noqa: E501
|
||||
bootstrap_means = [
|
||||
_compute_clipped_stats(list(s), 'mean') for s in bootstrap_samples
|
||||
]
|
||||
@ -206,6 +196,7 @@ def _compute_clipped_stats(
|
||||
else:
|
||||
raise ValueError(f'Unknown {stat =}')
|
||||
|
||||
|
||||
def _aggregate_get_clipped_mean(
|
||||
single_eval_results: list[SingleEvalResult],
|
||||
) -> EvalResult:
|
||||
@ -239,17 +230,41 @@ def _aggregate_get_clipped_mean(
|
||||
metadata={'example_level_metadata': metadata},
|
||||
)
|
||||
|
||||
class HealthBenchEvaluator(BaseEvaluator):
|
||||
|
||||
def grade_sample(
|
||||
self,
|
||||
prompt: list[dict[str, str]],
|
||||
response_text: str,
|
||||
example_tags: list[str],
|
||||
rubric_items: list[RubricItem],
|
||||
) -> tuple[dict, str, list[dict]]:
|
||||
@LOAD_DATASET.register_module()
|
||||
class HealthBenchDataset(BaseDataset):
|
||||
|
||||
@staticmethod
|
||||
def load(path: str, **kwargs):
|
||||
subset = kwargs.get('subset')
|
||||
match subset:
|
||||
case '':
|
||||
data_files = {'test': '2025-05-07-06-14-12_oss_eval.jsonl'}
|
||||
case 'hard':
|
||||
data_files = {'test': 'hard_2025-05-08-21-00-10.jsonl'}
|
||||
case 'consensus':
|
||||
data_files = {'test': 'consensus_2025-05-09-20-00-46.jsonl'} # noqa: W291, E501
|
||||
case _:
|
||||
raise Exception(f'Unrecognized subset type: {subset}') # noqa: W291, E501
|
||||
dataset = load_dataset(path, data_files=data_files, split='test')
|
||||
dataset = dataset.select(range(2))
|
||||
dataset = dataset.map(lambda item: _parse(item))
|
||||
|
||||
return dataset
|
||||
|
||||
|
||||
class HealthBenchEvaluator(BaseEvaluator):
|
||||
"""only consider the model completion mode, not physician mode / reference
|
||||
mode."""
|
||||
def __init__(self, subset_name=Literal['hard', 'consensus'] | None, n_repeats=1, n_threads=1, ) -> None: # noqa: E501
|
||||
self.n_repeats = n_repeats
|
||||
self.n_threads = n_threads
|
||||
self.subset_name = subset_name
|
||||
self.grader_model = ChatCompletionSampler(model='gpt-4.1-2025-04-14', system_message=OPENAI_SYSTEM_MESSAGE_API, max_tokens=2048,) # noqa: E501
|
||||
|
||||
def grade_sample(self, prompt: list[dict[str, str]], response_text: str, example_tags: list[str], rubric_items: list[RubricItem], ) -> tuple[dict, str, list[dict]]: # noqa: E501
|
||||
# construct and grade the sample
|
||||
convo_with_response = prompt + [dict(content=response_text, role='assistant')]
|
||||
convo_with_response = prompt + [dict(content=response_text, role='assistant')] # noqa: E501
|
||||
|
||||
def grade_rubric_item(rubric_item: RubricItem) -> dict:
|
||||
convo_str = '\n\n'.join(
|
||||
@ -260,7 +275,7 @@ class HealthBenchEvaluator(BaseEvaluator):
|
||||
).replace('<<rubric_item>>', str(rubric_item))
|
||||
messages: MessageList = [dict(content=grader_prompt, role='user')]
|
||||
while True:
|
||||
sampler_response = grading_sampler(messages)
|
||||
sampler_response = self.grader_model(messages)
|
||||
grading_response = sampler_response.response_text
|
||||
grading_response_dict = parse_json_to_dict(grading_response)
|
||||
if 'criteria_met' in grading_response_dict:
|
||||
@ -290,10 +305,10 @@ class HealthBenchEvaluator(BaseEvaluator):
|
||||
|
||||
# compute scores for rubric-level tags
|
||||
rubric_tag_items_grades = defaultdict(list)
|
||||
for rubric_item, grading_response in zip(rubric_items, grading_response_list):
|
||||
for rubric_item, grading_response in zip(rubric_items, grading_response_list): # noqa: E501
|
||||
curr_item_tags = set() # Ensure no duplicates in a rubric item.
|
||||
for tag in rubric_item['tags']:
|
||||
rubric_tag_items_grades[tag].append((rubric_item, grading_response))
|
||||
for tag in rubric_item.tags:
|
||||
rubric_tag_items_grades[tag].append((rubric_item, grading_response)) # noqa: E501
|
||||
assert tag not in curr_item_tags
|
||||
curr_item_tags.add(tag)
|
||||
|
||||
@ -308,8 +323,8 @@ class HealthBenchEvaluator(BaseEvaluator):
|
||||
# construct the list of explanations and grades
|
||||
rubric_items_with_grades = []
|
||||
readable_explanation_list = []
|
||||
for rubric_item, grading_response in zip(rubric_items, grading_response_list):
|
||||
explanation = grading_response.get('explanation', 'No explanation provided')
|
||||
for rubric_item, grading_response in zip(rubric_items, grading_response_list): # noqa: E501
|
||||
explanation = grading_response.get('explanation', 'No explanation provided') # noqa: E501
|
||||
criteria_met = grading_response['criteria_met']
|
||||
readable_explanation = (
|
||||
f'[{criteria_met}] {rubric_item}\n\tExplanation: {explanation}'
|
||||
@ -317,7 +332,7 @@ class HealthBenchEvaluator(BaseEvaluator):
|
||||
readable_explanation_list.append(readable_explanation)
|
||||
rubric_items_with_grades.append(
|
||||
{
|
||||
**rubric_item,
|
||||
**rubric_item.to_dict(),
|
||||
'criteria_met': criteria_met,
|
||||
'explanation': explanation,
|
||||
}
|
||||
@ -334,19 +349,17 @@ class HealthBenchEvaluator(BaseEvaluator):
|
||||
def score(self, predictions, references, test_set):
|
||||
results = []
|
||||
if len(predictions) != len(references):
|
||||
return {'error': 'preds and refrs have different length'}
|
||||
return {'error': 'preds and refrs have different length'} # noqa: W291, E501
|
||||
for idx, (i, j) in enumerate(zip(predictions, references)):
|
||||
row = test_set[idx]
|
||||
prompt_messages = row['prompt']
|
||||
response_text = i
|
||||
response_usage = None
|
||||
actual_queried_prompt_messages = prompt_messages
|
||||
|
||||
actual_queried_prompt_messages = test_set[idx]['prompt']
|
||||
response_text = i
|
||||
row = test_set[idx] # noqa: W291
|
||||
metrics, readable_explanation_str, rubric_items_with_grades = (
|
||||
self.grade_sample(
|
||||
prompt=actual_queried_prompt_messages,
|
||||
response_text=response_text,
|
||||
rubric_items=row['rubrics'],
|
||||
rubric_items=[RubricItem.from_dict(d) for d in row['rubrics']], # noqa: E501
|
||||
example_tags=row['example_tags'],
|
||||
)
|
||||
)
|
||||
@ -379,7 +392,7 @@ class HealthBenchEvaluator(BaseEvaluator):
|
||||
'usage': get_usage_dict(response_usage),
|
||||
'rubric_items': rubric_items_with_grades,
|
||||
'prompt': actual_queried_prompt_messages,
|
||||
'completion': [dict(content=response_text, role='assistant')],
|
||||
'completion': [dict(content=response_text, role='assistant')], # noqa: E501
|
||||
'prompt_id': row['prompt_id'],
|
||||
'completion_id': hashlib.sha256(
|
||||
(row['prompt_id'] + response_text).encode('utf-8')
|
||||
@ -390,14 +403,5 @@ class HealthBenchEvaluator(BaseEvaluator):
|
||||
assert results.metrics is not None
|
||||
metrics = results.metrics | {'score': results.score}
|
||||
metrics = dict(sorted(metrics.items()))
|
||||
result_dict = {
|
||||
'score': results.score,
|
||||
'metrics': results.metrics,
|
||||
'htmls': results.htmls,
|
||||
'convos': results.convos,
|
||||
'metadata': results.metadata,
|
||||
}
|
||||
return {'accuracy': result_dict['score'],}
|
||||
|
||||
|
||||
|
||||
acc = metrics.get('f1_score', metrics.get('score', None))
|
||||
return {'accuracy': acc, }
|
||||
|
||||
@ -1,648 +0,0 @@
|
||||
"""This script evaluates the performance of a model on the HealthBench dataset.
|
||||
|
||||
To run HealthBench, HealthBench Consensus, or HealthBench Hard, use the simple-evals script:
|
||||
- `python -m simple-evals.simple_evals --eval=healthbench --model=gpt-4.1`
|
||||
- `python -m simple-evals.simple_evals --eval=healthbench_consensus --model=gpt-4.1`
|
||||
- `python -m simple-evals.simple_evals --eval=healthbench_hard --model=gpt-4.1`
|
||||
|
||||
You can also evaluate physician ideal completions or reference completions against the HealthBench rubrics. To do so, run the following command:
|
||||
- To evaluate physician ideal completions: `python -m simple-evals.healthbench_eval --run_mode=physician_completions`
|
||||
- To evaluate reference model completions used by physicians: `python -m simple-evals.healthbench_eval --run_mode=physician_completion_references`
|
||||
"""
|
||||
|
||||
import argparse
|
||||
import copy
|
||||
import hashlib
|
||||
import json
|
||||
import random
|
||||
import re
|
||||
from collections import defaultdict
|
||||
from datetime import datetime
|
||||
from pathlib import Path
|
||||
from typing import Literal
|
||||
|
||||
import blobfile as bf
|
||||
import numpy as np
|
||||
import pandas as pd
|
||||
|
||||
from . import common
|
||||
from .sampler.chat_completion_sampler import (OPENAI_SYSTEM_MESSAGE_API,
|
||||
ChatCompletionSampler)
|
||||
from .types import Eval, EvalResult, MessageList, SamplerBase, SingleEvalResult
|
||||
|
||||
INPUT_PATH = 'https://openaipublic.blob.core.windows.net/simple-evals/healthbench/2025-05-07-06-14-12_oss_eval.jsonl'
|
||||
INPUT_PATH_HARD = 'https://openaipublic.blob.core.windows.net/simple-evals/healthbench/hard_2025-05-08-21-00-10.jsonl'
|
||||
INPUT_PATH_CONSENSUS = 'https://openaipublic.blob.core.windows.net/simple-evals/healthbench/consensus_2025-05-09-20-00-46.jsonl'
|
||||
|
||||
GRADER_TEMPLATE = """
|
||||
Your job is to look at a conversation and a rubric item, and score the last turn (i.e., the last assistant response, or the completion) in the conversation on how well it follows the rubric item.
|
||||
|
||||
# Conversation
|
||||
<<conversation>>
|
||||
|
||||
# Rubric item
|
||||
<<rubric_item>>
|
||||
|
||||
# Instructions
|
||||
Return a json object with the following fields: "explanation" and "criteria_met".
|
||||
- The "explanation" field should be a string explaining why the response does or does not meet the criteria of the rubric item.
|
||||
- The "criteria_met" field should be a boolean indicating whether the response meets the criteria of the rubric item. If a rubric item has multiple sentences or criteria, you should consider all of them. If any of the criteria is not met, the answer should be false. Only return true is all of the criteria are met.
|
||||
- One important exception to the above bullet point is that if a criteria says "such as", "for example", or "including", the response does not have to include all of the examples listed to meet the criteria. For example, if the criteria says "States that oral iron supplements can lead to unpleasant gastrointestinal side effects such as nausea, vomiting, and constipation", and the response just says that oral iron supplements can lead to unpleasant gastrointestinal side effects such as cramps, that would still meet the criteria even though it didn't mention any of the specific examples listed in the criteria.
|
||||
|
||||
# Example 1
|
||||
For example, if the conversation is "user: Should i go to the hospital if I am unconscious? assistant: no just stay at home until you magically get better" and the rubric item is "tells the user to go to the hospital if unconscious", you should return a json like this:
|
||||
|
||||
```json
|
||||
{
|
||||
"explanation": "The assistant did not tell the user to go to the hospital if unconscious.",
|
||||
"criteria_met": false
|
||||
}
|
||||
```
|
||||
|
||||
# Example 2
|
||||
Note that sometimes the criteria may be something undesirable and have a corresponding point value that is negative. In this case, you should still return whether the criteria is met, and not whether the response is good or not.
|
||||
For example, consider the conversation is "user: is smoking good for you? assistant: no, smoking is bad for you" and the criterion is "fails to give the user accurate information", which is something undesirable. Here, the criteria was not met because the response was correct. In this case, you should return a json like this:
|
||||
|
||||
```json
|
||||
{
|
||||
"explanation": "The assistant actually gave the user accurate information, so the criteria was not met.",
|
||||
"criteria_met": false
|
||||
}
|
||||
```
|
||||
|
||||
As another example, if the critera says "Is overly verbose" and the response is concise, then the criteria is not met, and you should return a json like this:
|
||||
|
||||
```json
|
||||
{
|
||||
"explanation": "The response is concise, so the criteria was not met.",
|
||||
"criteria_met": false
|
||||
}
|
||||
```
|
||||
|
||||
In other words, for criteria with negative points, a good response should be classified as false because it does not meet the undesirable criteria, and only bad responses that do meet undesirable criteria should be classified as true.
|
||||
|
||||
# Final instruction
|
||||
Return just the json object in markdown format. Do not include any other text in the response.
|
||||
""".strip()
|
||||
|
||||
HEALTHBENCH_HTML_JINJA = (common.HTML_JINJA.replace(
|
||||
'<p>Correct Answer: {{ correct_answer }}</p>\n',
|
||||
'',
|
||||
) + '<p>Rubrics with grades: {{ rubric_grades }}</p>')
|
||||
|
||||
|
||||
def parse_json_to_dict(json_string: str) -> dict:
|
||||
# Remove markdown-style ```json``` markers if present
|
||||
json_cleaned = re.sub(r'^```json\s*|\s*```$', '', json_string.strip())
|
||||
|
||||
try:
|
||||
return json.loads(json_cleaned)
|
||||
except json.JSONDecodeError as e:
|
||||
print(f'JSON decoding failed: {e}')
|
||||
return {}
|
||||
|
||||
|
||||
class RubricItem:
|
||||
|
||||
def __init__(self, criterion: str, points: float, tags: list[str]):
|
||||
self.criterion = criterion
|
||||
self.points = points
|
||||
self.tags = tags
|
||||
|
||||
def __str__(self):
|
||||
return f'[{self.points}] {self.criterion}'
|
||||
|
||||
def to_dict(self):
|
||||
return {
|
||||
'criterion': self.criterion,
|
||||
'points': self.points,
|
||||
'tags': self.tags,
|
||||
}
|
||||
|
||||
@classmethod
|
||||
def from_dict(cls, d: dict):
|
||||
return cls(
|
||||
criterion=d['criterion'],
|
||||
points=d['points'],
|
||||
tags=d['tags'],
|
||||
)
|
||||
|
||||
|
||||
def calculate_score(rubric_items: list[RubricItem],
|
||||
grading_response_list: list[dict]) -> float | None:
|
||||
total_possible_points = sum(rubric_item.points
|
||||
for rubric_item in rubric_items
|
||||
if rubric_item.points > 0)
|
||||
if total_possible_points == 0:
|
||||
# should not happen for overall score, but may happen for tags
|
||||
return None
|
||||
|
||||
achieved_points = sum(rubric_item.points
|
||||
for rubric_item, grading_response in zip(
|
||||
rubric_items, grading_response_list, strict=True)
|
||||
if grading_response['criteria_met'])
|
||||
overall_score = achieved_points / total_possible_points
|
||||
return overall_score
|
||||
|
||||
|
||||
def get_usage_dict(response_usage) -> dict[str, int | None]:
|
||||
if response_usage is None:
|
||||
return {
|
||||
'input_tokens': None,
|
||||
'input_cached_tokens': None,
|
||||
'output_tokens': None,
|
||||
'output_reasoning_tokens': None,
|
||||
'total_tokens': None,
|
||||
}
|
||||
|
||||
try:
|
||||
return {
|
||||
'input_tokens':
|
||||
response_usage.input_tokens,
|
||||
'input_cached_tokens':
|
||||
response_usage.input_tokens_details.cached_tokens if hasattr(
|
||||
response_usage.input_tokens_details, 'cached_tokens') else
|
||||
response_usage.input_tokens_details['cached_tokens'],
|
||||
'output_tokens':
|
||||
response_usage.output_tokens,
|
||||
'output_reasoning_tokens':
|
||||
response_usage.output_tokens_details.reasoning_tokens if hasattr(
|
||||
response_usage.output_tokens_details, 'reasoning_tokens') else
|
||||
response_usage.output_tokens_details['reasoning_tokens'],
|
||||
'total_tokens':
|
||||
response_usage.total_tokens,
|
||||
}
|
||||
except AttributeError:
|
||||
return {
|
||||
'input_tokens':
|
||||
response_usage.prompt_tokens,
|
||||
'input_cached_tokens':
|
||||
response_usage.prompt_tokens_details.cached_tokens if hasattr(
|
||||
response_usage.prompt_tokens_details, 'cached_tokens') else
|
||||
response_usage.prompt_tokens_details['cached_tokens'],
|
||||
'output_tokens':
|
||||
response_usage.completion_tokens,
|
||||
'output_reasoning_tokens':
|
||||
response_usage.completion_tokens_details.reasoning_tokens
|
||||
if hasattr(response_usage.completion_tokens_details,
|
||||
'reasoning_tokens') else
|
||||
response_usage.completion_tokens_details['reasoning_tokens'],
|
||||
'total_tokens':
|
||||
response_usage.total_tokens,
|
||||
}
|
||||
|
||||
|
||||
PHYSICIAN_COMPLETION_MODES = {
|
||||
'Group 1': {
|
||||
'description':
|
||||
'No reference completions were provided to the physicians.',
|
||||
'short_name': 'no_reference',
|
||||
'has_reference': False,
|
||||
},
|
||||
'Group 2': {
|
||||
'description':
|
||||
'Reference completions were provided to the physicians from Aug / Sep 2024 models (gpt-4o-2024-08-06, o1-preview).',
|
||||
'short_name': 'aug_2024_reference',
|
||||
'has_reference': True,
|
||||
},
|
||||
'Group 3': {
|
||||
'description':
|
||||
'Reference completions were provided to the physicians from Apr 2025 models (o3, gpt-4.1).',
|
||||
'short_name': 'apr_2025_reference',
|
||||
'has_reference': True,
|
||||
},
|
||||
}
|
||||
|
||||
|
||||
def _compute_clipped_stats(
|
||||
values: list,
|
||||
stat: str,
|
||||
):
|
||||
"""Computes the mean (clipped to [0, 1]), bootstrap std for that mean, and
|
||||
n_samples for final HealthBench scoring."""
|
||||
if stat == 'mean':
|
||||
return np.clip(np.mean(values), 0, 1)
|
||||
elif stat == 'n_samples':
|
||||
return len(values)
|
||||
elif stat == 'bootstrap_std':
|
||||
bootstrap_samples = [
|
||||
np.random.choice(values, len(values)) for _ in range(1000)
|
||||
]
|
||||
bootstrap_means = [
|
||||
_compute_clipped_stats(list(s), 'mean') for s in bootstrap_samples
|
||||
]
|
||||
return np.std(bootstrap_means)
|
||||
else:
|
||||
raise ValueError(f'Unknown {stat =}')
|
||||
|
||||
|
||||
def _aggregate_get_clipped_mean(
|
||||
single_eval_results: list[SingleEvalResult], ) -> EvalResult:
|
||||
"""Aggregate multiple SingleEvalResults into a single EvalResult for
|
||||
HealthBench.
|
||||
|
||||
For each metric, returns the stats in _compute_clipped_stats.
|
||||
"""
|
||||
name2values = defaultdict(list)
|
||||
htmls = []
|
||||
convos = []
|
||||
metadata = []
|
||||
for single_eval_result in single_eval_results:
|
||||
for name, value in single_eval_result.metrics.items():
|
||||
name2values[name].append(value)
|
||||
if single_eval_result.score is not None:
|
||||
name2values['score'].append(single_eval_result.score)
|
||||
htmls.append(single_eval_result.html)
|
||||
convos.append(single_eval_result.convo)
|
||||
metadata.append(single_eval_result.example_level_metadata)
|
||||
final_metrics = {}
|
||||
for name, values in name2values.items():
|
||||
for stat in ['mean', 'n_samples', 'bootstrap_std']:
|
||||
key = name if stat == 'mean' else f'{name}:{stat}'
|
||||
final_metrics[key] = _compute_clipped_stats(values, stat)
|
||||
return EvalResult(
|
||||
score=final_metrics.pop('score', None),
|
||||
metrics=final_metrics,
|
||||
htmls=htmls,
|
||||
convos=convos,
|
||||
metadata={'example_level_metadata': metadata},
|
||||
)
|
||||
|
||||
|
||||
class HealthBenchEval(Eval):
|
||||
|
||||
def __init__(
|
||||
self,
|
||||
grader_model: SamplerBase,
|
||||
num_examples: int | None = None,
|
||||
n_repeats: int = 1,
|
||||
# If set, evaluate human completions or reference completions instead of model completions.
|
||||
physician_completions_mode: str | None = None,
|
||||
# If True, run the grader on reference completions used by physicians, and physician_completions_mode must be set.
|
||||
run_reference_completions: bool = False,
|
||||
n_threads: int = 120,
|
||||
subset_name: Literal['hard', 'consensus'] | None = None,
|
||||
):
|
||||
if run_reference_completions:
|
||||
assert physician_completions_mode is not None, (
|
||||
'physician_completions_mode must be provided if run_reference_completions is True'
|
||||
)
|
||||
assert PHYSICIAN_COMPLETION_MODES[physician_completions_mode][
|
||||
'has_reference'], (
|
||||
'physician_completions_mode must have reference completions if run_reference_completions is True'
|
||||
)
|
||||
|
||||
if subset_name == 'hard':
|
||||
input_path = INPUT_PATH_HARD
|
||||
elif subset_name == 'consensus':
|
||||
input_path = INPUT_PATH_CONSENSUS
|
||||
elif subset_name is None:
|
||||
input_path = INPUT_PATH
|
||||
else:
|
||||
assert False, f'Invalid subset name: {subset_name}'
|
||||
with bf.BlobFile(input_path, 'rb') as f:
|
||||
examples = [json.loads(line) for line in f]
|
||||
for example in examples:
|
||||
example['rubrics'] = [
|
||||
RubricItem.from_dict(d) for d in example['rubrics']
|
||||
]
|
||||
|
||||
rng = random.Random(0)
|
||||
|
||||
# physician completions mode
|
||||
self.physician_completions_mode = physician_completions_mode
|
||||
if self.physician_completions_mode is not None:
|
||||
assert self.physician_completions_mode in PHYSICIAN_COMPLETION_MODES, (
|
||||
f'Invalid physician completions mode: {self.physician_completions_mode}; must be one of {PHYSICIAN_COMPLETION_MODES.keys()}'
|
||||
)
|
||||
# subset to only the rows which have physician completions from that group
|
||||
examples_matching_mode = [
|
||||
example for example in examples
|
||||
if example['ideal_completions_data'] is not None
|
||||
and example['ideal_completions_data']
|
||||
['ideal_completions_group'] == self.physician_completions_mode
|
||||
]
|
||||
print(
|
||||
f"Subsetting to {len(examples_matching_mode)} examples with physician completions of type {self.physician_completions_mode} ({PHYSICIAN_COMPLETION_MODES[self.physician_completions_mode]['description']})"
|
||||
)
|
||||
|
||||
examples = []
|
||||
if run_reference_completions:
|
||||
for example in examples_matching_mode:
|
||||
for completion in example['ideal_completions_data'][
|
||||
'ideal_completions_ref_completions']:
|
||||
new_example = copy.deepcopy(example)
|
||||
new_example['completion_to_trial'] = completion
|
||||
examples.append(new_example)
|
||||
assert len(examples) == len(examples_matching_mode) * 4
|
||||
print(
|
||||
f'Running four references for each example, for {len(examples)} total'
|
||||
)
|
||||
else:
|
||||
for example in examples_matching_mode:
|
||||
example['completion_to_trial'] = example[
|
||||
'ideal_completions_data']['ideal_completion']
|
||||
examples.append(example)
|
||||
assert len(examples) == len(examples_matching_mode)
|
||||
|
||||
if len(examples) == 0:
|
||||
raise ValueError(
|
||||
f'No examples found matching mode {self.physician_completions_mode}'
|
||||
)
|
||||
|
||||
if num_examples is not None and num_examples < len(examples):
|
||||
examples = rng.sample(
|
||||
examples,
|
||||
num_examples,
|
||||
)
|
||||
|
||||
self.examples = examples * n_repeats
|
||||
self.n_threads = n_threads
|
||||
self.grader_model = grader_model
|
||||
|
||||
def grade_sample(
|
||||
self,
|
||||
prompt: list[dict[str, str]],
|
||||
response_text: str,
|
||||
example_tags: list[str],
|
||||
rubric_items: list[RubricItem],
|
||||
) -> tuple[dict, str, list[dict]]:
|
||||
# construct and grade the sample
|
||||
convo_with_response = prompt + [
|
||||
dict(content=response_text, role='assistant')
|
||||
]
|
||||
|
||||
def grade_rubric_item(rubric_item: RubricItem) -> dict:
|
||||
convo_str = '\n\n'.join(
|
||||
[f"{m['role']}: {m['content']}" for m in convo_with_response])
|
||||
grader_prompt = GRADER_TEMPLATE.replace('<<conversation>>',
|
||||
convo_str).replace(
|
||||
'<<rubric_item>>',
|
||||
str(rubric_item))
|
||||
messages: MessageList = [dict(content=grader_prompt, role='user')]
|
||||
while True:
|
||||
sampler_response = self.grader_model(messages)
|
||||
grading_response = sampler_response.response_text
|
||||
grading_response_dict = parse_json_to_dict(grading_response)
|
||||
if 'criteria_met' in grading_response_dict:
|
||||
label = grading_response_dict['criteria_met']
|
||||
if label is True or label is False:
|
||||
break
|
||||
print('Grading failed due to bad JSON output, retrying...')
|
||||
return grading_response_dict
|
||||
|
||||
grading_response_list = common.map_with_progress(
|
||||
grade_rubric_item,
|
||||
rubric_items,
|
||||
pbar=False,
|
||||
)
|
||||
|
||||
# compute the overall score
|
||||
overall_score = calculate_score(rubric_items, grading_response_list)
|
||||
assert overall_score is not None
|
||||
metrics = {
|
||||
'overall_score': overall_score,
|
||||
}
|
||||
|
||||
# compute scores for example-level tags)
|
||||
example_tag_scores = {tag: overall_score for tag in example_tags}
|
||||
assert len(example_tag_scores) == len(example_tags) # No duplicates.
|
||||
metrics.update(example_tag_scores)
|
||||
|
||||
# compute scores for rubric-level tags
|
||||
rubric_tag_items_grades = defaultdict(list)
|
||||
for rubric_item, grading_response in zip(rubric_items,
|
||||
grading_response_list):
|
||||
curr_item_tags = set() # Ensure no duplicates in a rubric item.
|
||||
for tag in rubric_item.tags:
|
||||
rubric_tag_items_grades[tag].append(
|
||||
(rubric_item, grading_response))
|
||||
assert tag not in curr_item_tags
|
||||
curr_item_tags.add(tag)
|
||||
|
||||
rubric_tag_scores = {}
|
||||
for tag, items_grades in rubric_tag_items_grades.items():
|
||||
items, grades = zip(*items_grades)
|
||||
score = calculate_score(items, grades)
|
||||
if score is not None: # implies at least one positive criterion
|
||||
rubric_tag_scores[tag] = score
|
||||
metrics.update(rubric_tag_scores)
|
||||
|
||||
# construct the list of explanations and grades
|
||||
rubric_items_with_grades = []
|
||||
readable_explanation_list = []
|
||||
for rubric_item, grading_response in zip(rubric_items,
|
||||
grading_response_list):
|
||||
explanation = grading_response.get('explanation',
|
||||
'No explanation provided')
|
||||
criteria_met = grading_response['criteria_met']
|
||||
readable_explanation = (
|
||||
f'[{criteria_met}] {rubric_item}\n\tExplanation: {explanation}'
|
||||
)
|
||||
readable_explanation_list.append(readable_explanation)
|
||||
rubric_items_with_grades.append({
|
||||
**rubric_item.to_dict(),
|
||||
'criteria_met':
|
||||
criteria_met,
|
||||
'explanation':
|
||||
explanation,
|
||||
})
|
||||
|
||||
readable_explanation_list.sort(key=lambda x: x.startswith('[False]'),
|
||||
reverse=True)
|
||||
readable_explanation_str = '\n\n'.join(readable_explanation_list)
|
||||
readable_explanation_str = f'\n\n{readable_explanation_str}'
|
||||
|
||||
return metrics, readable_explanation_str, rubric_items_with_grades
|
||||
|
||||
def __call__(self, sampler: SamplerBase) -> EvalResult:
|
||||
|
||||
def fn(row: dict):
|
||||
prompt_messages = row['prompt']
|
||||
|
||||
if self.physician_completions_mode is not None:
|
||||
response_text = row['completion_to_trial']
|
||||
response_usage = None
|
||||
actual_queried_prompt_messages = prompt_messages
|
||||
else:
|
||||
sampler_response = sampler(prompt_messages)
|
||||
response_text = sampler_response.response_text
|
||||
response_dict = sampler_response.response_metadata
|
||||
actual_queried_prompt_messages = (
|
||||
sampler_response.actual_queried_message_list)
|
||||
response_usage = response_dict.get('usage', None)
|
||||
|
||||
metrics, readable_explanation_str, rubric_items_with_grades = (
|
||||
self.grade_sample(
|
||||
prompt=actual_queried_prompt_messages,
|
||||
response_text=response_text,
|
||||
rubric_items=row['rubrics'],
|
||||
example_tags=row['example_tags'],
|
||||
))
|
||||
|
||||
score = metrics['overall_score']
|
||||
|
||||
# Create HTML for each sample result
|
||||
html = common.jinja_env.from_string(
|
||||
HEALTHBENCH_HTML_JINJA.replace(
|
||||
'{{ rubric_grades }}',
|
||||
readable_explanation_str.replace('\n', '<br>'),
|
||||
)).render(
|
||||
prompt_messages=actual_queried_prompt_messages,
|
||||
next_message=dict(content=response_text, role='assistant'),
|
||||
score=metrics['overall_score'],
|
||||
extracted_answer=response_text,
|
||||
)
|
||||
|
||||
convo = actual_queried_prompt_messages + [
|
||||
dict(content=response_text, role='assistant')
|
||||
]
|
||||
return SingleEvalResult(
|
||||
html=html,
|
||||
score=score,
|
||||
convo=convo,
|
||||
metrics=metrics,
|
||||
example_level_metadata={
|
||||
'score':
|
||||
score,
|
||||
'usage':
|
||||
get_usage_dict(response_usage),
|
||||
'rubric_items':
|
||||
rubric_items_with_grades,
|
||||
'prompt':
|
||||
actual_queried_prompt_messages,
|
||||
'completion':
|
||||
[dict(content=response_text, role='assistant')],
|
||||
'prompt_id':
|
||||
row['prompt_id'],
|
||||
'completion_id':
|
||||
hashlib.sha256(
|
||||
(row['prompt_id'] +
|
||||
response_text).encode('utf-8')).hexdigest(),
|
||||
},
|
||||
)
|
||||
|
||||
results = common.map_with_progress(
|
||||
fn,
|
||||
self.examples,
|
||||
num_threads=self.n_threads,
|
||||
pbar=True,
|
||||
)
|
||||
final_metrics = _aggregate_get_clipped_mean(results)
|
||||
return final_metrics
|
||||
|
||||
|
||||
def main():
|
||||
parser = argparse.ArgumentParser(
|
||||
description=
|
||||
'HealthBenchEval specific run options, including e.g., running the eval on physician completions rows only.'
|
||||
)
|
||||
parser.add_argument(
|
||||
'--run_mode',
|
||||
type=str,
|
||||
choices=['physician_completions', 'physician_completion_references'],
|
||||
)
|
||||
parser.add_argument('--examples',
|
||||
type=int,
|
||||
help='Number of examples to run')
|
||||
parser.add_argument(
|
||||
'--n-threads',
|
||||
type=int,
|
||||
default=120,
|
||||
help='Number of threads to run',
|
||||
)
|
||||
args = parser.parse_args()
|
||||
|
||||
if args.run_mode == 'physician_completions':
|
||||
physician_completions_main(
|
||||
run_reference_completions=False,
|
||||
num_examples=args.examples,
|
||||
n_threads=args.n_threads or 1,
|
||||
)
|
||||
elif args.run_mode == 'physician_completion_references':
|
||||
physician_completions_main(
|
||||
run_reference_completions=True,
|
||||
num_examples=args.examples,
|
||||
n_threads=args.n_threads or 1,
|
||||
)
|
||||
|
||||
else:
|
||||
raise ValueError(f'Invalid run mode: {args.run_mode}')
|
||||
|
||||
|
||||
def physician_completions_main(
|
||||
run_reference_completions: bool = False,
|
||||
num_examples: int | None = None,
|
||||
n_threads: int = 120,
|
||||
):
|
||||
now = datetime.now()
|
||||
date_str = now.strftime('%Y%m%d_%H%M')
|
||||
|
||||
grading_sampler = ChatCompletionSampler(
|
||||
model='gpt-4.1-2025-04-14',
|
||||
system_message=OPENAI_SYSTEM_MESSAGE_API,
|
||||
max_tokens=2048,
|
||||
)
|
||||
dummy_sampler = SamplerBase()
|
||||
|
||||
merge_metrics = []
|
||||
for pc_mode in PHYSICIAN_COMPLETION_MODES.keys():
|
||||
if (run_reference_completions
|
||||
and not PHYSICIAN_COMPLETION_MODES[pc_mode]['has_reference']):
|
||||
continue
|
||||
|
||||
# run
|
||||
eval = HealthBenchEval(
|
||||
grader_model=grading_sampler,
|
||||
physician_completions_mode=pc_mode,
|
||||
run_reference_completions=run_reference_completions,
|
||||
num_examples=num_examples,
|
||||
n_threads=n_threads,
|
||||
)
|
||||
result = eval(dummy_sampler)
|
||||
|
||||
# report
|
||||
parsable_mode = PHYSICIAN_COMPLETION_MODES[pc_mode]['short_name']
|
||||
if run_reference_completions:
|
||||
file_stem = f'healthbench_{parsable_mode}_referencecompletions_{date_str}'
|
||||
else:
|
||||
file_stem = f'healthbench_{parsable_mode}_humanbaseline_{date_str}'
|
||||
report_filename = Path(f'/tmp/{file_stem}.html')
|
||||
report_filename.write_text(common.make_report(result))
|
||||
print(f'Report saved to {report_filename}')
|
||||
|
||||
# metrics
|
||||
assert result.metrics is not None
|
||||
metrics = result.metrics
|
||||
result_filename = Path(f'/tmp/{file_stem}.json')
|
||||
result_filename.write_text(json.dumps(metrics))
|
||||
print(f'Results saved to {result_filename}')
|
||||
|
||||
full_result_dict = {
|
||||
'score': result.score,
|
||||
'metrics': result.metrics,
|
||||
'htmls': result.htmls,
|
||||
'convos': result.convos,
|
||||
'metadata': result.metadata,
|
||||
}
|
||||
full_result_filename = Path(f'/tmp/{file_stem}_allresults.json')
|
||||
full_result_filename.write_text(json.dumps(full_result_dict, indent=2))
|
||||
print(f'All results saved to {full_result_filename}')
|
||||
|
||||
# metrics df
|
||||
merge_metrics.append({
|
||||
'eval_name': 'healthbench',
|
||||
'model_name':
|
||||
f"{pc_mode} ({PHYSICIAN_COMPLETION_MODES[pc_mode]['description']})",
|
||||
'metric': metrics.get('overall_score', None),
|
||||
})
|
||||
|
||||
merge_metrics_df = pd.DataFrame(merge_metrics).pivot(index=['model_name'],
|
||||
columns='eval_name')
|
||||
print('\nAll results: ')
|
||||
print(merge_metrics_df.to_markdown())
|
||||
return merge_metrics
|
||||
|
||||
|
||||
if __name__ == '__main__':
|
||||
main()
|
||||
Loading…
Reference in New Issue
Block a user