trace_synthesis/tasks/49.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

32 lines
918 B
JSON

{
"sites": [
"shopping"
],
"task_id": 49,
"require_login": true,
"storage_state": "./.auth/shopping_state.json",
"start_url": "http://localhost:28082",
"geolocation": null,
"intent_template": "Today is 6/12/2023. Tell me how many fulfilled orders I have {{period}}, and the total amount of money I spent.",
"instantiation_dict": {
"period": "over the past four month"
},
"intent": "Today is 6/12/2023. Tell me how many fulfilled orders I have over the past four month, and the total amount of money I spent.",
"require_reset": false,
"eval": {
"eval_types": [
"string_match"
],
"reference_answers": {
"fuzzy_match": [
"3 orders",
"$845.49 total spend"
]
},
"reference_url": "",
"program_html": [],
"string_note": "",
"reference_answer_raw_annotation": "3 orders, $845.49 total spend"
},
"intent_template_id": 197
}