trace_synthesis/tasks/34.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

37 lines
1.2 KiB
JSON

{
"sites": [
"map"
],
"task_id": 34,
"require_login": true,
"storage_state": null,
"start_url": "https://www.openstreetmap.org",
"geolocation": null,
"intent_template": "I will arrive {{place}} soon. Provide the name of a {{target1}} in the vicinity, if available. Then, tell me the {{information}} to {{target2}} from the hotel.",
"instantiation_dict": {
"place": "Pittsburgh Airport",
"target1": "Hyatt hotel",
"information": "the shortest walking time",
"target2": "a supermarket"
},
"intent": "I will arrive Pittsburgh Airport soon. Provide the name of a Hyatt hotel in the vicinity, if available. Then, tell me the the shortest walking time to a supermarket from the hotel.",
"require_reset": false,
"eval": {
"eval_types": [
"string_match"
],
"reference_answers": {
"must_include": [
"Hyatt Regency Pittsburgh International Airport"
],
"fuzzy_match": [
"Time: 3h 30min"
]
},
"reference_url": "",
"program_html": [],
"string_note": "",
"reference_answer_raw_annotation": "Hyatt Regency Pittsburgh International Airport\n3:30"
},
"intent_template_id": 78
}