[Update] Update dataset configs (#2030)

* [Update] Update dataset configs

* Fix lint

opencompass/configs/datasets/ChemBench/ChemBench_llmjudge_gen_c584cf.py

@@ -86,7 +86,7 @@ for _name in chembench_all_sets:
             ),
             dataset_cfg=dict(
                 type=ChemBenchDataset,
-                path='/fs-computility/llm/xiaolinchen/opencompass_fork/data/ChemBench4K',
+                path='opencompass/ChemBench4K',
                 name=_name,
                 reader_cfg=chembench_reader_cfg,
             ),

opencompass/datasets/TheoremQA/utils.py

@@ -33,7 +33,12 @@ def extract_theoremqa_answer(pred: str, answer_flag: bool = True):
             try:
                 with time_limit(1):
                     tmp = str(latex2sympy(pred))
-                    pred = str(eval(tmp))
+                    pred = eval(tmp)
+                    if isinstance(pred, tuple):
+                        pred = str(list(pred))
+                    else:
+                        pred = str(pred)
+
             except Exception:
                 if re.match(r'-?[\d\.]+\s\D+$', pred):
                     pred = pred.split(' ')[0]

opencompass/datasets/judge/rewardbench.py

@@ -14,6 +14,7 @@ from opencompass.utils import get_data_path
 
 from ..base import BaseDataset
 
+
 @LOAD_DATASET.register_module()
 class RewardBenchDataset(BaseDataset):
 

opencompass/datasets/smolinstruct.py

@@ -4,6 +4,7 @@ from collections import defaultdict
 
 import numpy as np
 from datasets import Dataset, DatasetDict, load_dataset
+from nltk.translate.meteor_score import meteor_score
 
 from opencompass.openicl.icl_evaluator.icl_base_evaluator import BaseEvaluator
 from opencompass.registry import (ICL_EVALUATORS, LOAD_DATASET,