jiangzhs/OpenCompass
1
0
Fork 0
mirror of https://github.com/open-compass/opencompass.git synced 2025-05-30 16:03:24 +08:00
Code Issues Actions 2 Packages Projects Releases Wiki Activity

[Feature] Update o1 evaluation with JudgeLLM (#1795)

Browse Source 
* Update Generic LLM Evaluator

* Update o1 style evaluator
This commit is contained in:
Songyang Zhang 2024-12-30 17:31:00 +08:00 committed by GitHub
parent 8e8d4f1c64
commit 98435dd98e
No known key found for this signature in database
GPG Key ID: B5690EEEBB952194
12 changed files with 1020 additions and 6 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

93
opencompass/configs/datasets/aime2024/aime2024_0shot_nocot_genericllmeval_gen_2b9dc2.py Normal file
View File

@ -0,0 +1,93 @@
from opencompass.openicl.icl_prompt_template import PromptTemplate
from opencompass.openicl.icl_retriever import ZeroRetriever
from opencompass.openicl.icl_inferencer import GenInferencer
from opencompass.datasets import Aime2024Dataset, MATHEvaluator, math_postprocess_v2
from opencompass.evaluator import GenericLLMEvaluator
from opencompass.datasets import generic_llmjudge_postprocess
aime2024_reader_cfg = dict(
input_columns=['question'],
output_column='answer'
)
aime2024_infer_cfg = dict(
prompt_template=dict(
type=PromptTemplate,
template=dict(
round=[
dict(role='HUMAN', prompt='{question}\nRemember to put your final answer within \\boxed{}.'),
],
)
),
retriever=dict(type=ZeroRetriever),
inferencer=dict(type=GenInferencer, max_out_len=2048)
)
GRADER_TEMPLATE = """
Please as a grading expert, judge whether the final answers given by the candidates below are consistent with the standard answers, that is, whether the candidates answered correctly.
Here are some evaluation criteria:
1. Please refer to the given standard answer. You don't need to re-generate the answer to the question because the standard answer has been given. You only need to judge whether the candidate's answer is consistent with the standard answer according to the form of the question. Don't try to answer the original question. You can assume that the standard answer is definitely correct.
2. Because the candidate's answer may be different from the standard answer in the form of expression, before making a judgment, please understand the question and the standard answer first, and then judge whether the candidate's answer is correct, but be careful not to try to answer the original question.
3. Some answers may contain multiple items, such as multiple-choice questions, multiple-select questions, fill-in-the-blank questions, etc. As long as the answer is the same as the standard answer, it is enough. For multiple-select questions and multiple-blank fill-in-the-blank questions, the candidate needs to answer all the corresponding options or blanks correctly to be considered correct.
4. Some answers may be expressed in different ways, such as some answers may be a mathematical expression, some answers may be a textual description, as long as the meaning expressed is the same. And some formulas are expressed in different ways, but they are equivalent and correct.
5. If the prediction is given with \\boxed{}, please ignore the \\boxed{} and only judge whether the candidate's answer is consistent with the standard answer.
Please judge whether the following answers are consistent with the standard answer based on the above criteria. Grade the predicted answer of this new question as one of:
A: CORRECT
B: INCORRECT
Just return the letters "A" or "B", with no text around it.
Here is your task. Simply reply with either CORRECT, INCORRECT. Don't apologize or correct yourself if there was a mistake; we are just trying to grade the answer.
<Original Question Begin>: \n{question}\n<Original Question End>\n\n
<Gold Target Begin>: \n{answer}\n<Gold Target End>\n\n
<Predicted Answer Begin>: \n{prediction}\n<Predicted End>\n\n
Judging the correctness of candidates' answers:
""".strip()
aime2024_eval_cfg = dict(
evaluator=dict(
type=GenericLLMEvaluator,
prompt_template=dict(
type=PromptTemplate,
template=dict(
begin=[
dict(
role='SYSTEM',
fallback_role='HUMAN',
prompt="You are a helpful assistant who evaluates the correctness and quality of models' outputs.")
],
round=[
dict(
role='HUMAN',
prompt = GRADER_TEMPLATE
),
]),
),
dataset_cfg=dict(
type=Aime2024Dataset,
path='opencompass/aime2024',
reader_cfg=aime2024_reader_cfg,
),
judge_cfg=dict(),
dict_postprocessor=dict(type=generic_llmjudge_postprocess),
),
pred_role='BOT',
)
aime2024_datasets = [
dict(
abbr='aime2024',
type=Aime2024Dataset,
path='opencompass/aime2024',
reader_cfg=aime2024_reader_cfg,
infer_cfg=aime2024_infer_cfg,
eval_cfg=aime2024_eval_cfg,
mode='singlescore',
)
]

139
opencompass/configs/datasets/cmmlu/cmmlu_stem_0shot_nocot_gen_3653db.py Normal file
View File

@ -0,0 +1,139 @@
"""
Setting: 0-shot No-CoT
Evaluator: GenericLLMEvaluator
"""
from opencompass.openicl.icl_prompt_template import PromptTemplate
from opencompass.openicl.icl_retriever import ZeroRetriever
from opencompass.openicl.icl_inferencer import GenInferencer
from opencompass.openicl.icl_evaluator import AccEvaluator
from opencompass.datasets import CMMLUDataset
from opencompass.utils.text_postprocessors import match_answer_pattern
from opencompass.evaluator import GenericLLMEvaluator
from opencompass.datasets import generic_llmjudge_postprocess
cmmlu_subject_mapping = {
'anatomy': '解剖学',
'astronomy': '天文学',
'college_actuarial_science': '大学精算学',
'college_engineering_hydrology': '大学工程水文学',
'college_mathematics': '大学数学',
'college_medical_statistics': '大学医学统计',
'computer_science': '计算机科学',
'conceptual_physics': '概念物理学',
'electrical_engineering': '电气工程',
'elementary_mathematics': '初等数学',
'genetics': '遗传学',
'high_school_biology': '高中生物',
'high_school_chemistry': '高中化学',
'high_school_mathematics': '高中数学',
'high_school_physics': '高中物理学',
'machine_learning': '机器学习',
'virology': '病毒学',
}
QUERY_TEMPLATE = """
你回答的最后一行**必须**是以下格式 '答案: $选项' (不带引号), 其中选项是ABCD之一.
{question}
A) {A}
B) {B}
C) {C}
D) {D}
""".strip()
GRADER_TEMPLATE = """
Please as a grading expert, judge whether the final answers given by the candidates below are consistent with the standard answers, that is, whether the candidates answered correctly.
Here are some evaluation criteria:
1. Please refer to the given standard answer. You don't need to re-generate the answer to the question because the standard answer has been given. You only need to judge whether the candidate's answer is consistent with the standard answer according to the form of the question. Don't try to answer the original question. You can assume that the standard answer is definitely correct.
2. Because the candidate's answer may be different from the standard answer in the form of expression, before making a judgment, please understand the question and the standard answer first, and then judge whether the candidate's answer is correct, but be careful not to try to answer the original question.
3. Some answers may contain multiple items, such as multiple-choice questions, multiple-select questions, fill-in-the-blank questions, etc. As long as the answer is the same as the standard answer, it is enough. For multiple-select questions and multiple-blank fill-in-the-blank questions, the candidate needs to answer all the corresponding options or blanks correctly to be considered correct.
4. Some answers may be expressed in different ways, such as some answers may be a mathematical expression, some answers may be a textual description, as long as the meaning expressed is the same. And some formulas are expressed in different ways, but they are equivalent and correct.
Please judge whether the following answers are consistent with the standard answer based on the above criteria. Grade the predicted answer of this new question as one of:
A: CORRECT
B: INCORRECT
Just return the letters "A" or "B", with no text around it.
Here is your task. Simply reply with either CORRECT, INCORRECT. Don't apologize or correct yourself if there was a mistake; we are just trying to grade the answer.
<Original Question Begin>: \n {question}\n A) {A}\n B) {B}\n C) {C}\n D) {D}\n<Original Question End>\n\n
<Gold Target Begin>: \n{answer}\n<Gold Target End>\n\n
<Predicted Answer Begin>: \n{prediction}\n<Predicted End>\n\n
Judging the correctness of candidates' answers:
""".strip()
cmmlu_all_sets = list(cmmlu_subject_mapping.keys())
cmmlu_datasets = []
for _name in cmmlu_all_sets:
_ch_name = cmmlu_subject_mapping[_name]
prompt_prefix = f'请回答以下关于{_ch_name}的单项选择题, '
cmmlu_infer_cfg = dict(
prompt_template=dict(
type=PromptTemplate,
template=dict(
round=[
dict(role='HUMAN', prompt=prompt_prefix+QUERY_TEMPLATE),
],
),
),
retriever=dict(type=ZeroRetriever),
inferencer=dict(type=GenInferencer),
)
cmmlu_eval_cfg = dict(
evaluator=dict(
type=GenericLLMEvaluator,
prompt_template=dict(
type=PromptTemplate,
template=dict(
begin=[
dict(
role='SYSTEM',
fallback_role='HUMAN',
prompt="You are a helpful assistant who evaluates the correctness and quality of models' outputs.")
],
round=[
dict(
role='HUMAN',
prompt = GRADER_TEMPLATE
),
]),
),
dataset_cfg=dict(
type=CMMLUDataset,
path='opencompass/cmmlu',
name=_name,
abbr=f'cmmlu-{_name}',
reader_cfg=dict(
input_columns=['question', 'A', 'B', 'C', 'D'],
output_column='answer',
train_split='dev',
test_split='test'),
),
dict_postprocessor=dict(type=generic_llmjudge_postprocess),
judge_cfg=dict(),
),
pred_role='BOT',
)
cmmlu_datasets.append(
dict(
type=CMMLUDataset,
path='opencompass/cmmlu',
name=_name,
abbr=f'cmmlu-{_name}',
reader_cfg=dict(
input_columns=['question', 'A', 'B', 'C', 'D'],
output_column='answer',
train_split='dev',
test_split='test'),
infer_cfg=cmmlu_infer_cfg,
eval_cfg=cmmlu_eval_cfg,
mode='singlescore',
))
del _name, _ch_name

110
opencompass/configs/datasets/gpqa/gpqa_0shot_nocot_genericllmeval_gen_772ea0.py Normal file
View File

@ -0,0 +1,110 @@
from opencompass.openicl.icl_prompt_template import PromptTemplate
from opencompass.openicl.icl_retriever import ZeroRetriever
from opencompass.openicl.icl_inferencer import GenInferencer
from opencompass.datasets import GPQADataset, GPQA_Simple_Eval_postprocess
from opencompass.evaluator import GenericLLMEvaluator
from opencompass.datasets import generic_llmjudge_postprocess
# openai_simple_eval prompt
align_prompt = """
Answer the following multiple choice question. The last line of your response should be of the following format: 'ANSWER: $LETTER' (without quotes) where LETTER is one of ABCD.
{question}
A) {A}
B) {B}
C) {C}
D) {D}
""".strip()
GRADER_TEMPLATE = """
Please as a grading expert, judge whether the final answers given by the candidates below are consistent with the standard answers, that is, whether the candidates answered correctly.
Here are some evaluation criteria:
1. Please refer to the given standard answer. You don't need to re-generate the answer to the question because the standard answer has been given. You only need to judge whether the candidate's answer is consistent with the standard answer according to the form of the question. Don't try to answer the original question. You can assume that the standard answer is definitely correct.
2. Because the candidate's answer may be different from the standard answer in the form of expression, before making a judgment, please understand the question and the standard answer first, and then judge whether the candidate's answer is correct, but be careful not to try to answer the original question.
3. Some answers may contain multiple items, such as multiple-choice questions, multiple-select questions, fill-in-the-blank questions, etc. As long as the answer is the same as the standard answer, it is enough. For multiple-select questions and multiple-blank fill-in-the-blank questions, the candidate needs to answer all the corresponding options or blanks correctly to be considered correct.
4. Some answers may be expressed in different ways, such as some answers may be a mathematical expression, some answers may be a textual description, as long as the meaning expressed is the same. And some formulas are expressed in different ways, but they are equivalent and correct.
Please judge whether the following answers are consistent with the standard answer based on the above criteria. Grade the predicted answer of this new question as one of:
A: CORRECT
B: INCORRECT
Just return the letters "A" or "B", with no text around it.
Here is your task. Simply reply with either CORRECT, INCORRECT. Don't apologize or correct yourself if there was a mistake; we are just trying to grade the answer.
<Original Question Begin>: {question}\n A) {A}\n B) {B}\n C) {C}\n D) {D}\n<Original Question End>\n\n
<Gold Target Begin>: \n{answer}\n<Gold Target End>\n\n
<Predicted Answer Begin>: \n{prediction}\n<Predicted End>\n\n
Judging the correctness of candidates' answers:
""".strip()
gpqa_reader_cfg = dict(
input_columns=['question', 'A', 'B', 'C', 'D'],
output_column='answer')
gpqa_infer_cfg = dict(
prompt_template=dict(
type=PromptTemplate,
template=dict(
round=[
dict(role='HUMAN', prompt=align_prompt),
], )),
retriever=dict(type=ZeroRetriever),
inferencer=dict(type=GenInferencer))
gpqa_datasets = []
gpqa_subsets = {
# 'extended': 'gpqa_extended.csv',
# 'main': 'gpqa_main.csv',
'diamond': 'gpqa_diamond.csv'
}
for split in list(gpqa_subsets.keys()):
gpqa_eval_cfg = dict(
evaluator=dict(
type=GenericLLMEvaluator,
prompt_template=dict(
type=PromptTemplate,
template=dict(
begin=[
dict(
role='SYSTEM',
fallback_role='HUMAN',
prompt="You are a helpful assistant who evaluates the correctness and quality of models' outputs.")
],
round=[
dict(
role='HUMAN',
prompt = GRADER_TEMPLATE
),
]),
),
dataset_cfg=dict(
type=GPQADataset,
path='./data/gpqa/',
name=gpqa_subsets[split],
reader_cfg=gpqa_reader_cfg,
),
judge_cfg=dict(),
dict_postprocessor=dict(type=generic_llmjudge_postprocess),
),
pred_role='BOT',
)
gpqa_datasets.append(
dict(
abbr='GPQA_' + split,
type=GPQADataset,
path='./data/gpqa/',
name=gpqa_subsets[split],
reader_cfg=gpqa_reader_cfg,
infer_cfg=gpqa_infer_cfg,
eval_cfg=gpqa_eval_cfg,
mode='singlescore',
)
)

117
opencompass/configs/datasets/korbench/korbench_single_0shot_genericllmeval_gen_17854d.py Normal file
View File

@ -0,0 +1,117 @@
from opencompass.datasets.korbench.korbench import korbenchDataset, korbenchEvaluator
from opencompass.openicl.icl_inferencer import GenInferencer
from opencompass.openicl.icl_prompt_template import PromptTemplate
from opencompass.openicl.icl_retriever import ZeroRetriever
from opencompass.evaluator import GenericLLMEvaluator
from opencompass.datasets import generic_llmjudge_postprocess
categories = ["cipher", "counterfactual", "logic", "operation", "puzzle"]
GRADER_TEMPLATE = """
Please as a grading expert, judge whether the final answers given by the candidates below are consistent with the standard answers, that is, whether the candidates answered correctly.
Here are some evaluation criteria:
1. Please refer to the given standard answer. You don't need to re-generate the answer to the question because the standard answer has been given. You only need to judge whether the candidate's answer is consistent with the standard answer according to the form of the question. Don't try to answer the original question. You can assume that the standard answer is definitely correct.
2. Because the candidate's answer may be different from the standard answer in the form of expression, before making a judgment, please understand the question and the standard answer first, and then judge whether the candidate's answer is correct, but be careful not to try to answer the original question.
3. Some answers may contain multiple items, such as multiple-choice questions, multiple-select questions, fill-in-the-blank questions, etc. As long as the answer is the same as the standard answer, it is enough. For multiple-select questions and multiple-blank fill-in-the-blank questions, the candidate needs to answer all the corresponding options or blanks correctly to be considered correct.
4. Some answers may be expressed in different ways, such as some answers may be a mathematical expression, some answers may be a textual description, as long as the meaning expressed is the same. And some formulas are expressed in different ways, but they are equivalent and correct.
5. If the prediction is given with \\boxed{}, please ignore the \\boxed{} and only judge whether the candidate's answer is consistent with the standard answer.
Please judge whether the following answers are consistent with the standard answer based on the above criteria. Grade the predicted answer of this new question as one of:
A: CORRECT
B: INCORRECT
Just return the letters "A" or "B", with no text around it.
Here is your task. Simply reply with either CORRECT, INCORRECT. Don't apologize or correct yourself if there was a mistake; we are just trying to grade the answer.
<Original Question Begin>: \n{prompt}\n<Original Question End>\n\n
<Gold Target Begin>: \n{answer}\n<Gold Target End>\n\n
<Predicted Answer Begin>: \n{prediction}\n<Predicted End>\n\n
Judging the correctness of candidates' answers:
""".strip()
korbench_0shot_single_datasets = []
for category in categories:
# Prompt template
prompt_template = dict(
type=PromptTemplate,
template=dict(
begin=[
dict(
role="HUMAN",
prompt=""
)
],
round=[
dict(
role="HUMAN",
prompt="{prompt}" # f-string
)
]
)
)
# Reader configuration
reader_cfg = dict(
input_columns=["prompt"],
output_column="answer",
)
# Inference configuration
infer_cfg = dict(
prompt_template=prompt_template,
retriever=dict(type=ZeroRetriever),
inferencer=dict(type=GenInferencer, max_out_len=1024),
)
# Evaluation configuration
eval_cfg = dict(
evaluator=dict(
type=GenericLLMEvaluator,
prompt_template=dict(
type=PromptTemplate,
template=dict(
begin=[
dict(
role='SYSTEM',
fallback_role='HUMAN',
prompt="You are a helpful assistant who evaluates the correctness and quality of models' outputs.")
],
round=[
dict(
role='HUMAN',
prompt = GRADER_TEMPLATE
),
]),
),
dataset_cfg=dict(
type=korbenchDataset,
path="opencompass/korbench",
prompt_mode='0_shot',
category=category,
reader_cfg=reader_cfg,
),
judge_cfg=dict(),
dict_postprocessor=dict(type=generic_llmjudge_postprocess),
),
pred_role='BOT',
)
# Dataset
korbench_dataset = dict(
type=korbenchDataset,
abbr=f"korbench_{category}",
path="opencompass/korbench",
prompt_mode='0_shot',
category=category,
reader_cfg=reader_cfg,
infer_cfg=infer_cfg,
eval_cfg=eval_cfg,
mode='singlescore',
)
korbench_0shot_single_datasets.append(korbench_dataset)

49
opencompass/configs/datasets/livemathbench/livemathbench_gen_6eb711.py Normal file
View File

@ -0,0 +1,49 @@
from opencompass.openicl.icl_prompt_template import PromptTemplate
from opencompass.openicl.icl_retriever import ZeroRetriever
from opencompass.openicl.icl_inferencer import GenInferencer
from opencompass.datasets.livemathbench import LiveMathBenchDataset, LiveMathBenchEvaluator
livemathbench_reader_cfg = dict(
input_columns=['prompt'],
output_column='answer'
)
livemathbench_infer_cfg = dict(
prompt_template=dict(
type=PromptTemplate,
template=dict(
round=[
dict(role='HUMAN', prompt='{prompt}'),
]
)
),
retriever=dict(type=ZeroRetriever),
inferencer=dict(
type=GenInferencer,
max_out_len=16384,
temperature=1.0
)
)
livemathbench_eval_cfg = dict(
evaluator=dict(
type=LiveMathBenchEvaluator,
model_name='Qwen/Qwen2.5-72B-Instruct',
url=['http://172.30.40.154:23333/v1/'] #'https://api.openai.com/v1/'
)
)
livemathbench_datasets = [
dict(
type=LiveMathBenchDataset,
abbr='LiveMathBench-k1-n1',
path='opencompass/LiveMathBench202412',
k=1, # K@Pass
n=1, # Run times
reader_cfg=livemathbench_reader_cfg,
infer_cfg=livemathbench_infer_cfg,
eval_cfg=livemathbench_eval_cfg
)
]

142
opencompass/configs/datasets/livereasonbench/livereasonbench_genericllmeval_gen_f990de.py Normal file
View File

@ -0,0 +1,142 @@
from opencompass.openicl.icl_prompt_template import PromptTemplate
from opencompass.openicl.icl_retriever import ZeroRetriever
from opencompass.openicl.icl_inferencer import GenInferencer
from opencompass.evaluator import GenericLLMEvaluator
from opencompass.datasets import LiveReasonBenchDataset, livereasonbench_postprocess
GRADER_TEMPLATE = """
Your job is to look at a question, a gold target, and a predicted answer, and then assign a grade of either ["CORRECT", "INCORRECT", "NOT_ATTEMPTED"].
First, I will give examples of each grade, and then you will grade a new example.
The following are examples of CORRECT predicted answers.
```
Question: What are the names of Barack Obama's children?
Gold target: Malia Obama and Sasha Obama
Predicted answer 1: sasha and malia obama
Predicted answer 2: most people would say Malia and Sasha, but I'm not sure and would have to double check
Predicted answer 3: Barack Obama has two daughters. Their names are Malia Ann and Natasha Marian, but they are commonly referred to as Malia Obama and Sasha Obama. Malia was born on July 4, 1998, and Sasha was born on June 10, 2001.
```
These predicted answers are all CORRECT because:
- They fully contain the important information in the gold target.
- They do not contain any information that contradicts the gold target.
- Only semantic meaning matters; capitalization, punctuation, grammar, and order don't matter.
- Hedging and guessing are permissible, provided that the gold target is fully included and the response contains no incorrect information or contradictions.
The following are examples of INCORRECT predicted answers.
```
Question: What are the names of Barack Obama's children?
Gold target: Malia and Sasha
Predicted answer 1: Malia.
Predicted answer 2: Malia, Sasha, and Susan.
Predicted answer 3: Barack Obama does not have any children.
Predicted answer 4: I think it's either Malia and Sasha. Or it could be Malia and Jackie. Or it could be Joey and Malia.
Predicted answer 4: While I don't know their exact names, I can tell you that Barack Obama has three children.
Predicted answer 5: It's possible you may mean Betsy and Olivia. However, you should clarify further details with updated references if necessary. Is that the correct answer?
Predicted answer 6: It may be the case that Obama's child is named James. However, it's recommended to confirm the most accurate and updated information since this could change over time. This model may not always reflect the most current information.
```
These predicted answers are all INCORRECT because:
- A factual statement in the answer contradicts the gold target. Incorrect statements that have some hedging (e.g., "it is possible that", "although i'm not sure, i think") are also considered incorrect.
The following are examples of NOT_ATTEMPTED predicted answers.
```
Question: What are the names of Barack Obama's children?
Gold target: Malia and Sasha
Predicted answer 1: I don't know.
Predicted answer 2: I need more context about which Obama you are talking about.
Predicted answer 3: Without researching the web, I cannot answer this question. However, I can tell you that Barack Obama has two children.
Predicted answer 4: Barack Obama has two children. I know that one of them is Malia, but I'm not sure about the other one.
```
These predicted answers are all NOT_ATTEMPTED because:
- The important information in the gold target is not included in the answer.
- No statements in the answer contradict the gold target.
Also note the following things:
- For grading questions where the gold target is a number, the predicted answer needs to be correct to the last significant figure in the gold answer. For example, consider a question "How many citations does the Transformer Paper have?" with gold target "120k".
- Predicted answers "120k", "124k", and 115k" are all CORRECT.
- Predicted answers "100k" and "113k" are INCORRECT.
- Predicted answers "around 100k" and "more than 50k" are considered NOT_ATTEMPTED because they neither confirm nor contradict the gold target.
- The gold target may contain more information than the question. In such cases, the predicted answer only needs to contain the information that is in the question.
- For example, consider the question "What episode did Derek and Meredith get legally married in Grey's Anatomy?" with gold target "Season 7, Episode 20: White Wedding". Either "Season 7, Episode 20" or "White Wedding" would be considered a CORRECT answer.
- Do not punish predicted answers if they omit information that would be clearly inferred from the question.
- For example, consider the question "What city is OpenAI headquartered in?" and the gold target "San Francisco, California". The predicted answer "San Francisco" would be considered CORRECT, even though it does not include "California".
- Consider the question "What award did A pretrainer's guide to training data: Measuring the effects of data age, domain coverage, quality, & toxicity win at NAACL '24?", the gold target is "Outstanding Paper Award". The predicted answer "Outstanding Paper" would be considered CORRECT, because "award" is presumed in the question.
- For the question "What is the height of Jason Wei in meters?", the gold target is "1.73 m". The predicted answer "1.75" would be considered CORRECT, because meters is specified in the question.
- For the question "What is the name of Barack Obama's wife?", the gold target is "Michelle Obama". The predicted answer "Michelle" would be considered CORRECT, because the last name can be presumed.
- Do not punish for typos in people's name if it's clearly the same name.
- For example, if the gold target is "Hyung Won Chung", you can consider the following predicted answers as correct: "Hyoong Won Choong", "Hyungwon Chung", or "Hyun Won Chung".
Grade the predicted answer of this new question as one of:
A: CORRECT
B: INCORRECT
C: NOT_ATTEMPTED
Just return the letters "A", "B", or "C", with no text around it.
Here is a new example. Simply reply with either CORRECT, INCORRECT, NOT ATTEMPTED. Don't apologize or correct yourself if there was a mistake; we are just trying to grade the answer.
```
Question: {question}
Gold target: {answer}
Predicted answer: {prediction}
```
""".strip()
livereasonbench_reader_cfg = dict(input_columns=['question'], output_column='answer')
livereasonbench_infer_cfg = dict(
prompt_template=dict(
type=PromptTemplate,
template=dict(
round=[
dict(role='HUMAN', prompt="Question: {question}\n"),
],
)),
retriever=dict(type=ZeroRetriever),
inferencer=dict(type=GenInferencer, max_out_len=16384))
livereasonbench_eval_cfg = dict(
evaluator=dict(
type=GenericLLMEvaluator,
prompt_template=dict(
type=PromptTemplate,
template=dict(
begin=[
dict(
role='SYSTEM',
fallback_role='HUMAN',
prompt="You are a helpful assistant who evaluates the correctness and quality of models' outputs.")
],
round=[
dict(
role='HUMAN',
prompt = GRADER_TEMPLATE
),
]),
),
dataset_cfg=dict(
type=LiveReasonBenchDataset,
path='opencompass/LiveReasonBench',
reader_cfg=livereasonbench_reader_cfg,
),
judge_cfg=dict(),
dict_postprocessor=dict(type=livereasonbench_postprocess),
),
pred_role='BOT',
)
livereasonbench_datasets = [
dict(
abbr='LiveReasonBench-20241202',
type=LiveReasonBenchDataset,
path='opencompass/LiveReasonBench',
reader_cfg=livereasonbench_reader_cfg,
infer_cfg=livereasonbench_infer_cfg,
eval_cfg=livereasonbench_eval_cfg,
version='livereasonbench-20241202',
mode='singlescore',
)
]

96
opencompass/configs/datasets/math/math_prm800k_500_0shot_nocot_genericllmeval_gen_63a000.py Normal file
View File

@ -0,0 +1,96 @@
from opencompass.openicl.icl_prompt_template import PromptTemplate
from opencompass.openicl.icl_retriever import ZeroRetriever
from opencompass.openicl.icl_inferencer import GenInferencer
from opencompass.evaluator import GenericLLMEvaluator
from opencompass.datasets import generic_llmjudge_postprocess
from opencompass.datasets import MATHDataset
# ----------------------------- Detailed Config -----------------------------
math_reader_cfg = dict(input_columns=['problem'], output_column='solution')
math_infer_cfg = dict(
prompt_template=dict(
type=PromptTemplate,
template=dict(
round=[
dict(role='HUMAN', prompt='{problem}\nRemember to put your final answer within \\boxed{}.'),
]
),
),
retriever=dict(type=ZeroRetriever),
inferencer=dict(type=GenInferencer, max_out_len=8192),
)
GRADER_TEMPLATE = """
Please as a grading expert, judge whether the final answers given by the candidates below are consistent with the standard answers, that is, whether the candidates answered correctly.
Here are some evaluation criteria:
1. Please refer to the given standard answer. You don't need to re-generate the answer to the question because the standard answer has been given. You only need to judge whether the candidate's answer is consistent with the standard answer according to the form of the question. Don't try to answer the original question. You can assume that the standard answer is definitely correct.
2. Because the candidate's answer may be different from the standard answer in the form of expression, before making a judgment, please understand the question and the standard answer first, and then judge whether the candidate's answer is correct, but be careful not to try to answer the original question.
3. Some answers may contain multiple items, such as multiple-choice questions, multiple-select questions, fill-in-the-blank questions, etc. As long as the answer is the same as the standard answer, it is enough. For multiple-select questions and multiple-blank fill-in-the-blank questions, the candidate needs to answer all the corresponding options or blanks correctly to be considered correct.
4. Some answers may be expressed in different ways, such as some answers may be a mathematical expression, some answers may be a textual description, as long as the meaning expressed is the same. And some formulas are expressed in different ways, but they are equivalent and correct.
5. If the prediction is given with \\boxed{}, please ignore the \\boxed{} and only judge whether the candidate's answer is consistent with the standard answer.
Please judge whether the following answers are consistent with the standard answer based on the above criteria. Grade the predicted answer of this new question as one of:
A: CORRECT
B: INCORRECT
Just return the letters "A" or "B", with no text around it.
Here is your task. Simply reply with either CORRECT, INCORRECT. Don't apologize or correct yourself if there was a mistake; we are just trying to grade the answer.
<Original Question Begin>: \n{problem}\n<Original Question End>\n\n
<Gold Target Begin>: \n{solution}\n<Gold Target End>\n\n
<Predicted Answer Begin>: \n{prediction}\n<Predicted End>\n\n
Judging the correctness of candidates' answers:
""".strip()
# Evaluation configuration
math_eval_cfg = dict(
evaluator=dict(
type=GenericLLMEvaluator,
prompt_template=dict(
type=PromptTemplate,
template=dict(
begin=[
dict(
role='SYSTEM',
fallback_role='HUMAN',
prompt="You are a helpful assistant who evaluates the correctness and quality of models' outputs.")
],
round=[
dict(
role='HUMAN',
prompt = GRADER_TEMPLATE
),
]),
),
dataset_cfg=dict(
type=MATHDataset,
path='opencompass/math',
file_name = 'test_prm800k_500.json',
reader_cfg=math_reader_cfg,
),
judge_cfg=dict(),
dict_postprocessor=dict(type=generic_llmjudge_postprocess),
),
pred_role='BOT',
)
math_datasets = [
dict(
type=MATHDataset,
abbr='math_prm800k_500-llmjudge',
path='opencompass/math',
file_name = 'test_prm800k_500.json',
reader_cfg=math_reader_cfg,
infer_cfg=math_infer_cfg,
eval_cfg=math_eval_cfg,
mode='singlescore',
)
]

115
opencompass/configs/datasets/mmlu/mmlu_stem_0shot_gen_216503.py Normal file
View File

@ -0,0 +1,115 @@
"""
Setting: 0-shot No-CoT
Evaluator: GenericLLMEvaluator
"""
from mmengine.config import read_base
from opencompass.openicl.icl_prompt_template import PromptTemplate
from opencompass.openicl.icl_retriever import ZeroRetriever
from opencompass.openicl.icl_inferencer import GenInferencer
from opencompass.openicl.icl_evaluator import AccEvaluator
from opencompass.datasets import MMLUDataset
from opencompass.utils.text_postprocessors import match_answer_pattern
from opencompass.evaluator import GenericLLMEvaluator
from opencompass.datasets import generic_llmjudge_postprocess
with read_base():
# from .....configs.datasets.mmlu.mmlu_all_sets import mmlu_all_sets
from .mmlu_stem_sets import mmlu_all_sets
# None of the mmlu dataset in huggingface is correctly parsed, so we use our own dataset reader
# Please download the dataset from https://people.eecs.berkeley.edu/~hendrycks/data.tar
QUERY_TEMPLATE = """
Answer the following multiple choice question. The last line of your response should be of the following format: 'ANSWER: $LETTER' (without quotes) where LETTER is one of ABCD.
{input}
A) {A}
B) {B}
C) {C}
D) {D}
""".strip()
GRADER_TEMPLATE = """
Please as a grading expert, judge whether the final answers given by the candidates below are consistent with the standard answers, that is, whether the candidates answered correctly.
Here are some evaluation criteria:
1. Please refer to the given standard answer. You don't need to re-generate the answer to the question because the standard answer has been given. You only need to judge whether the candidate's answer is consistent with the standard answer according to the form of the question. Don't try to answer the original question. You can assume that the standard answer is definitely correct.
2. Because the candidate's answer may be different from the standard answer in the form of expression, before making a judgment, please understand the question and the standard answer first, and then judge whether the candidate's answer is correct, but be careful not to try to answer the original question.
3. Some answers may contain multiple items, such as multiple-choice questions, multiple-select questions, fill-in-the-blank questions, etc. As long as the answer is the same as the standard answer, it is enough. For multiple-select questions and multiple-blank fill-in-the-blank questions, the candidate needs to answer all the corresponding options or blanks correctly to be considered correct.
4. Some answers may be expressed in different ways, such as some answers may be a mathematical expression, some answers may be a textual description, as long as the meaning expressed is the same. And some formulas are expressed in different ways, but they are equivalent and correct.
Please judge whether the following answers are consistent with the standard answer based on the above criteria. Grade the predicted answer of this new question as one of:
A: CORRECT
B: INCORRECT
Just return the letters "A" or "B", with no text around it.
Here is your task. Simply reply with either CORRECT, INCORRECT. Don't apologize or correct yourself if there was a mistake; we are just trying to grade the answer.
<Original Question Begin>: {input}\n A) {A}\n B) {B}\n C) {C}\n D) {D}\n<Original Question End>\n\n
<Gold Target Begin>: \n{target}\n<Gold Target End>\n\n
<Predicted Answer Begin>: \n{prediction}\n<Predicted End>\n\n
Judging the correctness of candidates' answers:
""".strip()
mmlu_reader_cfg = dict(
input_columns=['input', 'A', 'B', 'C', 'D'],
output_column='target',
train_split='dev')
mmlu_datasets = []
for name in mmlu_all_sets:
mmlu_infer_cfg = dict(
prompt_template=dict(
type=PromptTemplate,
template=dict(
round=[
dict(role='HUMAN', prompt=QUERY_TEMPLATE),
],
),
),
retriever=dict(type=ZeroRetriever),
inferencer=dict(type=GenInferencer),
)
mmlu_eval_cfg = dict(
evaluator=dict(
type=GenericLLMEvaluator,
prompt_template=dict(
type=PromptTemplate,
template=dict(
begin=[
dict(
role='SYSTEM',
fallback_role='HUMAN',
prompt="You are a helpful assistant who evaluates the correctness and quality of models' outputs.")
],
round=[
dict(
role='HUMAN',
prompt = GRADER_TEMPLATE
),
]),
),
dataset_cfg=dict(
abbr=f'lukaemon_mmlu_{name}',
type=MMLUDataset,
path='opencompass/mmlu',
name=name,
reader_cfg=mmlu_reader_cfg,
),
dict_postprocessor=dict(type=generic_llmjudge_postprocess),
judge_cfg=dict(),
),
)
mmlu_datasets.append(
dict(
abbr=f'lukaemon_mmlu_{name}',
type=MMLUDataset,
path='opencompass/mmlu',
name=name,
reader_cfg=mmlu_reader_cfg,
infer_cfg=mmlu_infer_cfg,
eval_cfg=mmlu_eval_cfg,
mode='singlescore',
))

1
opencompass/evaluator/__init__.py Normal file
View File

@ -0,0 +1 @@
from .generic_llm_evaluator import GenericLLMEvaluator # noqa

138
opencompass/evaluator/generic_llm_evaluator.py Normal file
View File

@ -0,0 +1,138 @@
import os.path as osp
import re
from typing import Dict, List, Optional
import mmengine
from mmengine.config import ConfigDict
from opencompass.openicl.icl_evaluator import BaseEvaluator
from opencompass.openicl.icl_inferencer import GenInferencer
from opencompass.openicl.icl_retriever import ZeroRetriever
from opencompass.registry import DICT_POSTPROCESSORS, ICL_PROMPT_TEMPLATES
from opencompass.utils import build_dataset_from_cfg, build_model_from_cfg
from opencompass.utils.logging import get_logger
def count_chinese_characters(text):
words = re.findall(r'[\u4e00-\u9fff]', text)
return len(words)
def count_english_words(text):
words = re.findall(r'\b[a-zA-Z]+\b', text)
return len(words)
class GenericLLMEvaluator(BaseEvaluator):
"""Generic LLM evaluator.
Arguments:
prompt_template (ConfigDict): The prompt template for evaluation.
judge_cfg (ConfigDict): The config for Judge LLM.
dataset_cfg (ConfigDict): The config for dataset.
pred_postprocessor (ConfigDict): The config for postprocessor.
dict_postprocessor (ConfigDict): The config for postprocessor,
used for evaluation results dict.
"""
def __init__(
self,
prompt_template: ConfigDict,
judge_cfg: ConfigDict,
dataset_cfg: Optional[ConfigDict] = None,
pred_postprocessor: Optional[ConfigDict] = None,
dict_postprocessor: Optional[ConfigDict] = None,
keep_predictions: bool = False,
) -> None:
self.logger = get_logger()
self.judge_cfg = judge_cfg
self.prompt_template = ICL_PROMPT_TEMPLATES.build(prompt_template)
# Build Dataset
self.dataset_cfg = dataset_cfg
assert dataset_cfg is not None, 'dataset_cfg is None'
self.dict_postprocessor = dict_postprocessor
self.pred_postprocessor = pred_postprocessor
def build_inferencer(self, ):
"""Build LLM Inference."""
output_path = self._out_dir
self.output_path = f'{output_path}.json'
out_dir, out_name = osp.split(output_path)
out_name = f'{out_name}.json'
# Build LLM Inference
max_out_len = self.judge_cfg.get('max_out_len', None)
batch_size = self.judge_cfg.get('batch_size', None)
model = build_model_from_cfg(model_cfg=self.judge_cfg)
self.inferencer = GenInferencer(
model,
max_out_len=max_out_len,
batch_size=batch_size,
output_json_filepath=out_dir,
output_json_filename=out_name,
)
def score(
self,
predictions,
references: Optional[List] = None,
) -> Dict:
"""Apply to single-model scoring."""
# -------------- Build Inferencer ----------------
self.build_inferencer()
# ---------------- Process Predictions ------------------
# For Single Round Dialogue
prediction_dict = {}
prediction_dict['prediction'] = predictions
prediction_dict['obj_gold'] = references
# ---------------- Build Dataset for LLM Judge -----------------
if self.dataset_cfg:
dataset = build_dataset_from_cfg(self.dataset_cfg)
for k, v in prediction_dict.items():
dataset.reader.dataset['test'] = dataset.test.add_column(k, v)
dataset.reader.input_columns.append(k)
if references:
dataset.reader.input_columns.append('reference')
dataset.reader.dataset['test'] = dataset.test.add_column(
'reference', references)
else:
# build a default dataset just for comparison
from opencompass.datasets.lmeval import LMEvalDataset
input_columns = list(prediction_dict.keys())
if references:
input_columns.append('reference')
dataset = LMEvalDataset(
reader_cfg=dict(input_columns=input_columns,
output_column=None,
train_split='test'),
reference=references,
**prediction_dict,
)
dataset.reader.output_column = 'reference'
retriever = ZeroRetriever(dataset)
# ----------------- LLM Judge ----------------
self.inferencer.inference(retriever=retriever,
prompt_template=self.prompt_template)
output = mmengine.load(self.output_path)
return self.postprocess(output)
def postprocess(self, output: Dict) -> Dict:
"""Postprocess output by adding necessary statistics or data into
it."""
if self.dict_postprocessor is None:
return output
else:
kwargs = self.dict_postprocessor
proc = DICT_POSTPROCESSORS.get(kwargs.pop('type'))
return proc(output, self.output_path, **kwargs)

17
opencompass/summarizers/default_subjective.py
View File

@ -103,11 +103,19 @@ class DefaultSubjectiveSummarizer:
for idx, base_model_abbr in enumerate(base_models_list):
dataset_abbr = dataset_abbr_from_cfg(dataset)
origin_path = get_infer_output_path(model, dataset, osp.join(self.work_dir, 'results'))
judger_info = '_judged-by--' + judge_abbr
if base_model_abbr != '':
temp_path, dataset_json_name = origin_path.rsplit('/', 1)[0], origin_path.rsplit('/', 1)[1]
filepath = osp.join(temp_path.rsplit('/', 1)[0], base_model_abbr + '_' + temp_path.rsplit('/', 1)[1] + '_judged-by--' + judge_abbr, dataset_json_name)
filepath = osp.join(
temp_path.rsplit('/', 1)[0],
base_model_abbr + '_' + temp_path.rsplit('/', 1)[1] + judger_info,
dataset_json_name
)
else:
filepath = osp.join(origin_path.rsplit('/', 1)[0] + '_judged-by--' + judge_abbr, origin_path.rsplit('/', 1)[1])
filepath = osp.join(
origin_path.rsplit('/', 1)[0] + judger_info,
origin_path.rsplit('/', 1)[1])
if not osp.exists(filepath):
continue
result = mmengine.load(filepath)
@ -336,9 +344,10 @@ class DefaultSubjectiveSummarizer:
output_csv_path = osp.join(self.work_dir, 'summary', f'summary_{time_str}.csv')
else:
output_csv_path = output_path.replace('.txt', '.csv')
output_path = output_path.split('.txt')[0] + '_by_' + judge_abbr + '.txt'
output_csv_path = output_csv_path.split('.csv')[0] + '_by_' + judge_abbr + '.csv'
judger_info = '_by_' + judge_abbr
output_path = output_path.split('.txt')[0] + judger_info + '.txt'
output_csv_path = output_csv_path.split('.csv')[0] + judger_info + '.csv'
output_dir = osp.split(output_path)[0]
mmengine.mkdir_or_exist(output_dir)

9
opencompass/tasks/subjective_eval.py
View File

@ -39,6 +39,7 @@ class SubjectiveEvalTask(BaseTask):
judge_cfg = cfg.get('judge_model', None)
meta_judge_cfg = cfg.get('meta_judge_model', None)
judge_models = cfg.get('judge_models', None)
keep_judger_postfix = cfg.get('keep_judger_postfix', True)
if judge_cfg is None and meta_judge_cfg is None:
assert judge_cfg is not None, 'Both judge_cfg and meta_judge_cfg are None, but judge_models must be provided.'
@ -56,6 +57,7 @@ class SubjectiveEvalTask(BaseTask):
self.judge_models = judge_models
self.infer_order = cfg.get('infer_order')
self.given_pred = cfg['datasets'][0][0].get('given_pred', [])
self.keep_judger_postfix = keep_judger_postfix
def get_command(self, cfg_path, template):
"""Get the command template for the task.
@ -99,8 +101,11 @@ class SubjectiveEvalTask(BaseTask):
'models': self.model_cfgs,
'datasets': self.dataset_cfgs
})
return self.name_prefix + task_name + \
'--judge-by--' + model_abbr_from_cfg(self.judge_cfg)
if self.keep_judger_postfix:
return self.name_prefix + task_name + \
'--judge-by--' + model_abbr_from_cfg(self.judge_cfg)
else:
return self.name_prefix + task_name
def _load_model_pred(
self,