trace_synthesis/tasks/81.json
yuyr a84d51a101 1. 增加r1生成综合策略代码和输出;
2. 增加tasks;
3. 增加analysis部分,对策略进行归纳分类,然后进行评测。
2025-04-17 17:40:15 +08:00

33 lines
948 B
JSON

{
"sites": [
"map"
],
"task_id": 81,
"require_login": true,
"storage_state": null,
"start_url": "https://www.openstreetmap.org",
"geolocation": null,
"intent_template": "What is the duration required to first walk from {{place_A}} to {{place_B}}, and then drive to {{place_C}}?",
"instantiation_dict": {
"place_A": "Univ of Pittsburgh",
"place_B": "starbucks on Craig Street",
"place_C": "Pittsburgh International Airport"
},
"intent": "What is the duration required to first walk from Univ of Pittsburgh to starbucks on Craig Street, and then drive to Pittsburgh International Airport?",
"require_reset": false,
"eval": {
"eval_types": [
"string_match"
],
"reference_answers": {
"fuzzy_match": [
"49 min"
]
},
"reference_url": "",
"program_html": [],
"string_note": "",
"reference_answer_raw_annotation": "49 min"
},
"intent_template_id": 72
}