OpenCompass/configs/datasets/tydiqa/tydiqa_gen_978d2a.py

62 lines
3.5 KiB
Python
Raw Normal View History

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 TydiQADataset, TydiQAEvaluator
[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 os import environ
# All configs are for TydiQA Goldp task
tydiqa_reader_cfg = dict(
2024-05-14 15:35:58 +08:00
input_columns=['passage_text', 'question_text'],
output_column='answer'
)
langs = ['arabic', 'bengali', 'english', 'finnish', 'indonesian', 'japanese', 'korean', 'russian', 'swahili', 'telugu', 'thai']
prefixs_prompt = {
2024-05-14 15:35:58 +08:00
'english': ('Answer the following question based on the information in the given passage.', 'Passage:', 'Question:', 'Answer:'),
'arabic': ('أجب على السؤال التالي بناءً على المعلومات في المقطع المعطى.', 'المقطع:', 'السؤال:', 'الإجابة:'),
'bengali': ('প্রদত্ত অধ্যায়ের তথ্যের উপর ভিত্তি করে নিম্নলিখিত প্রশ্নের উত্তর দিন।', 'অধ্যায়:', 'প্রশ্ন:', 'উত্তর:'),
'finnish': ('Vastaa seuraavaan kysymykseen annetun kappaleen tiedon perusteella.', 'Kappale:', 'Kysymys:', 'Vastaus:'),
'indonesian': ('Jawab pertanyaan berikut berdasarkan informasi di bagian yang diberikan.', 'Bagian:', 'Pertanyaan:', 'Jawaban:'),
'korean': ('주어진 문단의 정보에 기반하여 다음 질문에 답하십시오.', '문단:', '질문:', '답변:'),
'japanese':('文脈に基づいて質問に答えてください。','ぶんしょう:','しつもん:', 'かいとう:'),
'russian': ('Ответьте на следующий вопрос на основе информации в данном отрывке.', 'Отрывок:', 'Вопрос:', 'Ответ:'),
'swahili': ('Jibu swali lifuatalo kulingana na habari kwenye kifungu kilichotolewa.', 'Kifungu:', 'Swali:', 'Jibu:'),
'telugu': ('ఇచ్చిన పేరాలోని సమాచారం ఆధారంగా కింది ప్రశ్నకు సమాధానం ఇవ్వండి.', 'పేరా:', 'ప్రశ్న:', 'సమాధానం:'),
'thai':('ตอบคำถามต่อไปนี้โดยอิงตามข้อมูลในตอนข้อความที่กำหนด:', 'ตอนข้อความ:', 'คำถาม:', 'คำตอบ:')
}
tydiqa_datasets = []
for _lang in langs:
_hint = prefixs_prompt[_lang]
tydiqa_infer_cfg = dict(
prompt_template=dict(
type=PromptTemplate,
2024-05-14 15:35:58 +08:00
template=f'{_hint[0]}\n\n</E>{_hint[1]}{{passage_text}}\n{_hint[2]} {{question_text}}\n{_hint[3]} {{answer}}' ,
ice_token='</E>'
),
retriever=dict(type=ZeroRetriever),
inferencer=dict(type=GenInferencer), max_out_len=50
)
tydiqa_eval_cfg = dict(
evaluator=dict(type=TydiQAEvaluator),
ds_split='validation',
ds_column='answer',
)
[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
# Skip japanese due to filter rules of Modelscope
if environ.get('DATASET_SOURCE') == 'Modelscope' and _lang == 'japanese':
continue
tydiqa_datasets.append(
2023-11-15 15:23:58 +08:00
dict(abbr=f'tydiqa-goldp_{_lang}',
type=TydiQADataset,
[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='opencompass/tydiqa',
lang=_lang,
reader_cfg=tydiqa_reader_cfg,
infer_cfg=tydiqa_infer_cfg,
eval_cfg=tydiqa_eval_cfg
)
)