tag:
  - longbench
task: longbench_triviaqa
dataset_path: THUDM/LongBench
test_split: test
dataset_name: triviaqa
doc_to_text: "Answer the question based on the given passage. Only give me the answer and do not output any other words. The following are some examples.\n\n{{context}}\n\n{{input}}"
doc_to_target: '{{answers}}'
process_results: !function metrics.get_qa_f1_score
generation_kwargs:
  max_gen_toks: 32
  temperature: 1
  do_sample: False
  until: ["\n"]
metric_list:
  - metric: "qa_f1_score"
    aggregation: mean
    higher_is_better: True
metadata:
  version: 4.0