OpenCompass/configs/datasets/gpqa/gpqa_openai_simple_evals_gen_5aeece.py
Songyang Zhang 704853e5e7
[Feature] Update pip install (#1324)
* [Feature] Update pip install

* Update Configuration

* Update

* Update

* Update

* Update Internal Config

* Update collect env
2024-07-29 18:32:50 +08:00

53 lines
1.6 KiB
Python

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, GPQAEvaluator
# 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. Think step by step before answering.
{question}
A) {A}
B) {B}
C) {C}
D) {D}
""".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_eval_cfg = dict(evaluator=dict(type=GPQAEvaluator),
pred_postprocessor=dict(type=GPQA_Simple_Eval_postprocess))
gpqa_datasets = []
gpqa_subsets = {
# 'extended': 'gpqa_extended.csv',
# 'main': 'gpqa_main.csv',
'diamond': 'gpqa_diamond.csv'
}
for split in list(gpqa_subsets.keys()):
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)
)