trace_synthesis/tasks/711.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

44 lines
1.2 KiB
JSON

{
"sites": [
"shopping_admin"
],
"task_id": 711,
"require_login": true,
"storage_state": "./.auth/shopping_admin_state.json",
"start_url": "http://localhost:28083/admin",
"geolocation": null,
"intent_template": "Create a {{type}} report from {{start_date}} to {{end_date}}",
"instantiation_dict": {
"type": "product view",
"start_date": "07/05/2021",
"end_date": "05/31/2023"
},
"intent": "Create a product view report from 07/05/2021 to 05/31/2023",
"require_reset": false,
"eval": {
"eval_types": [
"url_match",
"program_html"
],
"reference_answers": null,
"reference_url": "http://localhost:28083/admin/reports/report_product/viewed/",
"program_html": [
{
"url": "last",
"locator": "document.querySelector('[id=\"sales_report_from\"').value",
"required_contents": {
"exact_match": "7/5/21"
}
},
{
"url": "last",
"locator": "document.querySelector('[id=\"sales_report_to\"').value",
"required_contents": {
"exact_match": "5/31/23"
}
}
],
"url_note": "GOLD in PRED"
},
"intent_template_id": 271
}