mirror of
https://github.com/open-compass/opencompass.git
synced 2025-05-30 16:03:24 +08:00
edab1c07ba
* 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>
89 lines
3.2 KiB
Python
89 lines
3.2 KiB
Python
import json
|
|
import os
|
|
|
|
from datasets import Dataset, DatasetDict
|
|
|
|
from opencompass.registry import LOAD_DATASET
|
|
from opencompass.utils import get_data_path
|
|
|
|
from .base import BaseDataset
|
|
|
|
|
|
@LOAD_DATASET.register_module()
|
|
class RoleBenchBaseDataset(BaseDataset):
|
|
|
|
@staticmethod
|
|
def load_single(source_file, desc_list):
|
|
with open(source_file, 'r', encoding='utf-8') as f:
|
|
source_data = [json.loads(line) for line in f.readlines()]
|
|
dataset = [{
|
|
'role': item['role'],
|
|
'desc': desc_list[item['role']],
|
|
'question': item['question'],
|
|
'answer': item['generated'][0]
|
|
} for item in source_data]
|
|
return dataset
|
|
|
|
@staticmethod
|
|
def load_desc(path):
|
|
path = get_data_path(path, local_mode=True)
|
|
with open(path, 'r', encoding='utf-8') as f:
|
|
desc_list = json.load(f)
|
|
return desc_list
|
|
|
|
@staticmethod
|
|
def load_dataset(path, desc_list):
|
|
train_data_list = RoleBenchBaseDataset.load_single(
|
|
os.path.join(path, 'general/train.jsonl'), desc_list)
|
|
train_data_list.extend(
|
|
RoleBenchBaseDataset.load_single(
|
|
os.path.join(path, 'role_specific/train.jsonl'), desc_list))
|
|
test_dataset = RoleBenchBaseDataset.load_single(
|
|
os.path.join(path, 'general/test.jsonl'), desc_list)
|
|
test_dataset.extend(
|
|
RoleBenchBaseDataset.load_single(
|
|
os.path.join(path, 'role_specific/test.jsonl'), desc_list))
|
|
return Dataset.from_list(train_data_list).shuffle(
|
|
seed=42), Dataset.from_list(test_dataset).shuffle(seed=42)
|
|
|
|
|
|
@LOAD_DATASET.register_module()
|
|
class InstructionGeneralizationEnglishDataset(RoleBenchBaseDataset):
|
|
|
|
@staticmethod
|
|
def load(path):
|
|
path = get_data_path(path, local_mode=True)
|
|
desc_list = RoleBenchBaseDataset.load_desc(
|
|
os.path.join(path, 'profiles-eng/desc.json'))
|
|
path = os.path.join(path, 'rolebench-eng/instruction-generalization')
|
|
train_dataset, test_dataset = RoleBenchBaseDataset.load_dataset(
|
|
path, desc_list)
|
|
return DatasetDict({'train': train_dataset, 'test': test_dataset})
|
|
|
|
|
|
@LOAD_DATASET.register_module()
|
|
class RoleGeneralizationEnglishDataset(RoleBenchBaseDataset):
|
|
|
|
@staticmethod
|
|
def load(path):
|
|
path = get_data_path(path, local_mode=True)
|
|
desc_list = RoleBenchBaseDataset.load_desc(
|
|
os.path.join(path, 'profiles-eng/desc.json'))
|
|
path = os.path.join(path, 'rolebench-eng/role-generalization')
|
|
train_dataset, test_dataset = RoleBenchBaseDataset.load_dataset(
|
|
path, desc_list)
|
|
return DatasetDict({'train': train_dataset, 'test': test_dataset})
|
|
|
|
|
|
@LOAD_DATASET.register_module()
|
|
class InstructionGeneralizationChineseDataset(RoleBenchBaseDataset):
|
|
|
|
@staticmethod
|
|
def load(path):
|
|
desc_list = RoleBenchBaseDataset.load_desc(
|
|
os.path.join(path, 'profiles-zh/desc.json'))
|
|
path = os.path.join(path, 'rolebench-zh')
|
|
train_dataset, test_dataset = RoleBenchBaseDataset.load_dataset(
|
|
path, desc_list)
|
|
return DatasetDict({'train': train_dataset, 'test': test_dataset})
|