Merge branch 'main' into mathvista

# Conflicts:
#	lm_eval/models/openai_completions.py

lm_eval/tasks/kbl/bar_exam/criminal/kbl_bar_exam_em_criminal_2014.yaml

+task: kbl_bar_exam_em_criminal_2014
+dataset_name: bar_exam_criminal_2014
+include: _base_em_yaml

lm_eval/tasks/kbl/bar_exam/criminal/kbl_bar_exam_em_criminal_2015.yaml

+task: kbl_bar_exam_em_criminal_2015
+dataset_name: bar_exam_criminal_2015
+include: _base_em_yaml

lm_eval/tasks/kbl/bar_exam/criminal/kbl_bar_exam_em_criminal_2016.yaml

+task: kbl_bar_exam_em_criminal_2016
+dataset_name: bar_exam_criminal_2016
+include: _base_em_yaml

lm_eval/tasks/kbl/bar_exam/criminal/kbl_bar_exam_em_criminal_2017.yaml

+task: kbl_bar_exam_em_criminal_2017
+dataset_name: bar_exam_criminal_2017
+include: _base_em_yaml

lm_eval/tasks/kbl/bar_exam/criminal/kbl_bar_exam_em_criminal_2018.yaml

+task: kbl_bar_exam_em_criminal_2018
+dataset_name: bar_exam_criminal_2018
+include: _base_em_yaml

lm_eval/tasks/kbl/bar_exam/criminal/kbl_bar_exam_em_criminal_2019.yaml

+task: kbl_bar_exam_em_criminal_2019
+dataset_name: bar_exam_criminal_2019
+include: _base_em_yaml

lm_eval/tasks/kbl/bar_exam/criminal/kbl_bar_exam_em_criminal_2020.yaml

+task: kbl_bar_exam_em_criminal_2020
+dataset_name: bar_exam_criminal_2020
+include: _base_em_yaml

lm_eval/tasks/kbl/bar_exam/criminal/kbl_bar_exam_em_criminal_2021.yaml

+task: kbl_bar_exam_em_criminal_2021
+dataset_name: bar_exam_criminal_2021
+include: _base_em_yaml

lm_eval/tasks/kbl/bar_exam/criminal/kbl_bar_exam_em_criminal_2022.yaml

+task: kbl_bar_exam_em_criminal_2022
+dataset_name: bar_exam_criminal_2022
+include: _base_em_yaml

lm_eval/tasks/kbl/bar_exam/criminal/kbl_bar_exam_em_criminal_2023.yaml

+task: kbl_bar_exam_em_criminal_2023
+dataset_name: bar_exam_criminal_2023
+include: _base_em_yaml

lm_eval/tasks/kbl/bar_exam/criminal/kbl_bar_exam_em_criminal_2024.yaml

+task: kbl_bar_exam_em_criminal_2024
+dataset_name: bar_exam_criminal_2024
+include: _base_em_yaml

lm_eval/tasks/kbl/bar_exam/public/_base_em_yaml

+tag:
+  - kbl
+  - kbl_bar_exam_em
+  - kbl_bar_exam_em_public
+description: '당신은 사용자의 질문에 친절하고 논리적으로 답변해 주는 법률 전문가 챗봇 입니다.\n'
+dataset_path: lbox/kbl
+test_split: test
+output_type: generate_until
+doc_to_text: '### 질문: {{question}}
+
+  다음 각 선택지를 읽고 A, B, C, D, E 중 하나를 선택하여 ''답변: A'' 와 같이 단답식으로 답해 주세요.
+
+  A. {{A}}
+
+  B. {{B}}
+
+  C. {{C}}
+
+  D. {{D}}
+
+  E. {{E}}
+
+  ### 답변:'
+doc_to_target: gt
+metric_list:
+  - metric: exact_match
+    aggregation: mean
+    higher_is_better: true
+    ignore_case: true
+    ignore_punctuation: true
+filter_list:
+  - name: get-answer
+    filter:
+    - function: regex
+      regex_pattern: ([A-E]).*
+    - function: take_first

lm_eval/tasks/kbl/bar_exam/public/kbl_bar_exam_em_public_2012.yaml

+task: kbl_bar_exam_em_public_2012
+dataset_name: bar_exam_public_2012
+include: _base_em_yaml

lm_eval/tasks/kbl/bar_exam/public/kbl_bar_exam_em_public_2013.yaml

+task: kbl_bar_exam_em_public_2013
+dataset_name: bar_exam_public_2013
+include: _base_em_yaml

lm_eval/tasks/kbl/bar_exam/public/kbl_bar_exam_em_public_2014.yaml

+task: kbl_bar_exam_em_public_2014
+dataset_name: bar_exam_public_2014
+include: _base_em_yaml

lm_eval/tasks/kbl/bar_exam/public/kbl_bar_exam_em_public_2015.yaml

+task: kbl_bar_exam_em_public_2015
+dataset_name: bar_exam_public_2015
+include: _base_em_yaml

lm_eval/tasks/kbl/bar_exam/public/kbl_bar_exam_em_public_2016.yaml

+task: kbl_bar_exam_em_public_2016
+dataset_name: bar_exam_public_2016
+include: _base_em_yaml

lm_eval/tasks/kbl/bar_exam/public/kbl_bar_exam_em_public_2017.yaml

+task: kbl_bar_exam_em_public_2017
+dataset_name: bar_exam_public_2017
+include: _base_em_yaml

lm_eval/tasks/kbl/bar_exam/public/kbl_bar_exam_em_public_2018.yaml

+task: kbl_bar_exam_em_public_2018
+dataset_name: bar_exam_public_2018
+include: _base_em_yaml

lm_eval/tasks/kbl/bar_exam/public/kbl_bar_exam_em_public_2019.yaml

+task: kbl_bar_exam_em_public_2019
+dataset_name: bar_exam_public_2019
+include: _base_em_yaml