trace_synthesis/tasks/284.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

27 lines
889 B
JSON

{
"sites": [
"shopping"
],
"task_id": 284,
"require_login": true,
"storage_state": "./.auth/shopping_state.json",
"start_url": "http://localhost:28082",
"geolocation": null,
"intent_template": "Show the least expensive {{product}} with a minimum storage capacity of {{min_storage}}.",
"instantiation_dict": {
"product": "shoe storage",
"min_storage": "12 pairs"
},
"intent": "Show the least expensive shoe storage with a minimum storage capacity of 12 pairs.",
"require_reset": false,
"eval": {
"eval_types": [
"url_match"
],
"reference_answers": null,
"reference_url": "http://localhost:28082/onlyeasy-over-the-door-shoe-storage-organizer-hanging-shoe-rack-holder-with-24-large-fabric-pockets-22-1-x-61-4-herringbone-grey-mxrodsb1p.html",
"program_html": [],
"url_note": "GOLD in PRED"
},
"intent_template_id": 207
}