1{2 "document_id": "lb_qasper_88d89e5b02c860bd1fdac17796e2b6048a6d2b86950c4c12",3 "family": "public_qasper",4 "split": "test",5 "document_hash": "5bd4f773836dc599be000ef2fdee880c32c7351828a8f7bf5c55b01d4bdbdd4a",6 "questions": [7 {8 "question_id": "q00",9 "qtype": "qasper",10 "question": "Which existing benchmarks did they compare to?",11 "answer": "Affective Text, Fairy Tales, ISEAR",12 "aliases": [13 " Affective Text dataset, Fairy Tales dataset, ISEAR dataset"14 ],15 "answer_type": "qa_f1",16 "tolerance": 0.0,17 "supporting_fact_ids": [],18 "answer_position_frac": -1.019 }20 ],21 "facts": []22}