dataset_path: Idavidrein/gpqa
tag: gpqa
output_type: generate_until
process_docs: !function utils.process_docs
training_split: train
# Because huggingface dataset only has train split
validation_split: train
test_split: null
doc_to_text: "What is the correct answer to this question:{{Question}}\nChoices:\n(A) {{choice1}}\n(B) {{choice2}}\n(C) {{choice3}}\n(D) {{choice4}}\nPlease reason step by step, and present your final answer using the format: Final Answer: \\boxed{X}, where X is one of A, B, C, or D."
doc_to_target: answer
filter_list:
  - name: "strict-match"
    filter:
      - function: "regex"
        regex_pattern: "(?<=Final Answer: )\\boxed{[A-D]}"
      - function: "take_first"
  - name: "flexible-extract"
    filter:
      - function: "multi_choice_regex"
        group_select: -1
        ignore_case: true
        ignore_punctuation: true
        regex_pattern: "\\{[A-Z]\\}"
      - function: "take_first"
generation_kwargs:
  until: []
  do_sample: true
  temperature: 0.6
  top_p: 0.95
  max_gen_toks: 16384
num_fewshot: 0
metric_list:
  - metric: exact_match
    aggregation: mean
    higher_is_better: true
    ignore_case: true
    ignore_punctuation: true
metadata:
  version: 1.0
