✨ feat(eval): add external scoring mode (#12729)

* wip: add llm relevant & BrowseComp

* wip: add widesearch desc

* wip: dsqa, hle, widesearch

* wip: add dsqa

* wip: add awaiting eval status for runs

* wip: add awaiting status for run

* wip: adjust hle-verified

* 🐛 fix: browsecomp topics

* 📝 docs: add annotations

* wip: add awaiting status for pass@k

* wip: add complete status

* wip: update theard dots

* wip: update run status page

* wip: remove useless impl

* wip: update prompt

* ✨ feat: add external eval routes

* wip: add eval cli

* 🐛 fix: support authoritize in no browser environment

* wip: pass tests

* ♻️ refactor: remove tests

* ♻️ refactor: mo camel case

locales/en-US/eval.json

@@ -157,13 +157,15 @@
   "difficulty.easy": "Easy",
   "difficulty.hard": "Hard",
   "difficulty.medium": "Medium",
+  "evalMode.answer-relevance": "LLM Relevance",
+  "evalMode.answer-relevance.desc": "Use LLM to evaluate answer relevance (yes or no)",
   "evalMode.contains": "Contains Match",
   "evalMode.contains.desc": "Output must contain the expected text",
   "evalMode.equals": "Exact Match",
   "evalMode.equals.desc": "Output must be exactly the same as expected",
   "evalMode.label": "Eval Mode",
   "evalMode.llm-rubric": "LLM Judge",
-  "evalMode.llm-rubric.desc": "Use LLM to evaluate output quality",
+  "evalMode.llm-rubric.desc": "Use LLM to evaluate output quality based on custom criteria (0.0 to 1.0)",
   "evalMode.placeholder": "Select eval mode",
   "evalMode.prompt.label": "Judge Prompt",
   "evalMode.prompt.placeholder": "Enter the evaluation criteria or prompt for LLM judge",
@@ -256,12 +258,16 @@
   "run.running.hint": "Evaluation is running, results will appear shortly...",
   "run.status.aborted": "Aborted",
   "run.status.completed": "Completed",
+  "run.status.completed.tooltip": "This evaluation has completed running all test cases and scoring.",
   "run.status.error": "Run Error",
+  "run.status.external": "External",
+  "run.status.external.tooltip": "This evaluation is waiting for external scoring. Results will be updated when scoring is complete.",
   "run.status.failed": "Failed",
   "run.status.idle": "Idle",
   "run.status.pending": "Pending",
   "run.status.running": "Running",
   "run.status.timeout": "Timeout",
+  "sidebar": "Evaluation",
   "sidebar.benchmarks": "Benchmarks",
   "sidebar.dashboard": "Dashboard",
   "sidebar.datasets": "Datasets",

locales/zh-CN/eval.json

@@ -161,6 +161,8 @@
   "evalMode.contains.desc": "输出中必须包含期望的文本",
   "evalMode.equals": "精确匹配",
   "evalMode.equals.desc": "输出必须与期望内容完全一致",
+  "evalMode.external": "外部评估",
+  "evalMode.external.desc": "智能体完成运行后，由外部系统提交评估结果",
   "evalMode.label": "评估模式",
   "evalMode.llm-rubric": "LLM 评判",
   "evalMode.llm-rubric.desc": "使用 LLM 评估输出质量",
@@ -256,7 +258,10 @@
   "run.running.hint": "评测进行中，结果即将呈现...",
   "run.status.aborted": "已终止",
   "run.status.completed": "已完成",
+  "run.status.completed.tooltip": "评测已完成运行，所有结果已评估。",
   "run.status.error": "运行出错",
+  "run.status.external": "待外部评测",
+  "run.status.external.tooltip": "智能体已完成运行，等待外部系统提交评估结果。",
   "run.status.failed": "失败",
   "run.status.idle": "待开始",
   "run.status.pending": "等待中",