[Update] Update Math auto-download data (#1700)

This commit is contained in:
Linchen Xiao 2024-11-18 20:24:35 +08:00 committed by GitHub
parent 98242ff1d1
commit ab8fdbbaab
No known key found for this signature in database
GPG Key ID: B5690EEEBB952194
2 changed files with 10 additions and 8 deletions

View File

@ -1,4 +1,5 @@
import json
import os
import re
from os import environ
@ -140,7 +141,7 @@ def extract_answer(response_text: str):
class MATHDataset(BaseDataset):
@staticmethod
def load(path: str):
def load(path: str, file_name: str = 'math.json'):
path = get_data_path(path)
dataset = DatasetDict()
raw_data = []
@ -155,7 +156,8 @@ class MATHDataset(BaseDataset):
extract_boxed_answer(item['solution'])
})
else:
data = json.load(open(path))
file_path = os.path.join(path, file_name)
data = json.load(open(file_path))
for i in data.keys():
raw_data.append({
'problem':

View File

@ -167,7 +167,7 @@ DATASETS_MAPPING = {
"opencompass/math": {
"ms_id": "opencompass/math",
"hf_id": "opencompass/math",
"local": "./data/math/math.json",
"local": "./data/math/",
},
# MMLU
"opencompass/mmlu": {
@ -327,11 +327,11 @@ DATASETS_MAPPING = {
"hf_id": "",
"local": "./data/mmmlu_lite",
},
"opencompass/musr": {
"opencompass/musr": {
"ms_id": "",
"hf_id": "",
"local": "./data/musr",
},
},
"opencompass/babilong": {
"ms_id": "",
"hf_id": "",
@ -374,7 +374,7 @@ DATASETS_URL = {
},
"/math/": {
"url": "http://opencompass.oss-cn-shanghai.aliyuncs.com/datasets/data/math.zip",
"md5": "8b1b897259684672055e6fd4fc07c808",
"md5": "cb5b4c8378085929e20345174e731fdf",
},
"/hellaswag/": {
"url": "http://opencompass.oss-cn-shanghai.aliyuncs.com/datasets/data/hellaswag.zip",
@ -540,8 +540,8 @@ DATASETS_URL = {
"url": "http://opencompass.oss-cn-shanghai.aliyuncs.com/datasets/data/WikiBench.zip",
"md5": "6dac1d1a3133fe1effff185cbf71d928",
},
"/babilong":{
"/babilong": {
"url": "http://opencompass.oss-cn-shanghai.aliyuncs.com/datasets/data/babilong.zip",
"md5": "e400864c31bc58d29eaa3e199751f99b",
}
},
}