OpenCompass/opencompass/datasets/wikibench.py

65 lines
2.2 KiB
Python
Raw Normal View History

import copy
import json
from datasets import Dataset
from opencompass.registry import LOAD_DATASET
[Feature] Support ModelScope datasets (#1289) * add ceval, gsm8k modelscope surpport * update race, mmlu, arc, cmmlu, commonsenseqa, humaneval and unittest * update bbh, flores, obqa, siqa, storycloze, summedits, winogrande, xsum datasets * format file * format file * update dataset format * support ms_dataset * udpate dataset for modelscope support * merge myl_dev and update test_ms_dataset * udpate dataset for modelscope support * update readme * update eval_api_zhipu_v2 * remove unused code * add get_data_path function * update readme * remove tydiqa japanese subset * add ceval, gsm8k modelscope surpport * update race, mmlu, arc, cmmlu, commonsenseqa, humaneval and unittest * update bbh, flores, obqa, siqa, storycloze, summedits, winogrande, xsum datasets * format file * format file * update dataset format * support ms_dataset * udpate dataset for modelscope support * merge myl_dev and update test_ms_dataset * update readme * udpate dataset for modelscope support * update eval_api_zhipu_v2 * remove unused code * add get_data_path function * remove tydiqa japanese subset * update util * remove .DS_Store * fix md format * move util into package * update docs/get_started.md * restore eval_api_zhipu_v2.py, add environment setting * Update dataset * Update * Update * Update * Update --------- Co-authored-by: Yun lin <yunlin@U-Q9X2K4QV-1904.local> Co-authored-by: Yunnglin <mao.looper@qq.com> Co-authored-by: Yun lin <yunlin@laptop.local> Co-authored-by: Yunnglin <maoyl@smail.nju.edu.cn> Co-authored-by: zhangsongyang <zhangsongyang@pjlab.org.cn>
2024-07-29 13:48:32 +08:00
from opencompass.utils import get_data_path
from .base import BaseDataset
def get_number(options):
result_string = ''
for i, option in enumerate(options, start=65):
result_string += f'{chr(i)}. {option}\n'
return result_string
@LOAD_DATASET.register_module()
class WikiBenchDataset(BaseDataset):
@staticmethod
def load(path: str, name: str):
[Feature] Support ModelScope datasets (#1289) * add ceval, gsm8k modelscope surpport * update race, mmlu, arc, cmmlu, commonsenseqa, humaneval and unittest * update bbh, flores, obqa, siqa, storycloze, summedits, winogrande, xsum datasets * format file * format file * update dataset format * support ms_dataset * udpate dataset for modelscope support * merge myl_dev and update test_ms_dataset * udpate dataset for modelscope support * update readme * update eval_api_zhipu_v2 * remove unused code * add get_data_path function * update readme * remove tydiqa japanese subset * add ceval, gsm8k modelscope surpport * update race, mmlu, arc, cmmlu, commonsenseqa, humaneval and unittest * update bbh, flores, obqa, siqa, storycloze, summedits, winogrande, xsum datasets * format file * format file * update dataset format * support ms_dataset * udpate dataset for modelscope support * merge myl_dev and update test_ms_dataset * update readme * udpate dataset for modelscope support * update eval_api_zhipu_v2 * remove unused code * add get_data_path function * remove tydiqa japanese subset * update util * remove .DS_Store * fix md format * move util into package * update docs/get_started.md * restore eval_api_zhipu_v2.py, add environment setting * Update dataset * Update * Update * Update * Update --------- Co-authored-by: Yun lin <yunlin@U-Q9X2K4QV-1904.local> Co-authored-by: Yunnglin <mao.looper@qq.com> Co-authored-by: Yun lin <yunlin@laptop.local> Co-authored-by: Yunnglin <maoyl@smail.nju.edu.cn> Co-authored-by: zhangsongyang <zhangsongyang@pjlab.org.cn>
2024-07-29 13:48:32 +08:00
path = get_data_path(path, local_mode=True)
circular_patterns = ['ABCD', 'BCDA', 'CDAB', 'DABC']
data = []
2023-12-11 17:42:53 +08:00
with open(path, 'r', encoding='utf-8') as infile:
for id, line in enumerate(infile):
entry = json.loads(line)
if 'cloze' in name:
data.append({
'question': entry['question'].strip(),
'answer': entry['answer'].strip()
})
elif 'circular' in name:
for c in circular_patterns:
line = copy.deepcopy(entry)
options = []
for i in range(4):
options.append(line['options'][ord(c[i]) -
ord('A')])
line['options'] = options
line['answer'] = {
c[0]: 'A',
c[1]: 'B',
c[2]: 'C',
c[3]: 'D'
}[line['answer']]
line['answer'] = str(
id) + '--' + line['answer'] + '--' + c
line['question'] = line['question'].strip(
) + '\n' + get_number(line['options'])
data.append(line)
else:
# treat as normal single choice question
entry['question'] = entry['question'].strip(
) + '\n' + get_number(entry['options'])
data.append(entry)
dataset = Dataset.from_list(data)
return dataset