Files
ZacharyZhang-NY 77219c930b Add third-party benchmark suite frozen at 50,789 sources / 5,308 queries
Ground truth comes only from third-party annotations (MMLongBench-Doc
evidence_page_ids, VisRAG qrels, ScreenSpot-v2 bboxes, Video-MME QA,
LongMemEval answer_session_ids, CodeSearchNet func_path+func_name) and
from document structure; no model judgement is used anywhere.

- bench/corpus/manifest.jsonl: 28 sources with url, retrieved_at, sha256,
  annotation_source, license, files[]; verified by bench.manifest_check
- bench/sets/{pdf,image,ui_screenshot,video,code,conversation}.{jsonl} plus
  corpus indexes and a manifest that freezes every file's sha256
- 2,928 answerable + 2,380 unanswerable queries over 854 holders; every
  query carries scope {holder, scenario, as_of} for transactive routing
- bench/validate_sets.py fails on missing gold, holder mismatch between
  gold and query, gold observed after as_of, and manifest digest drift
2026-10-06 12:05:43 -04:00

74 lines
2.1 KiB
JSON

{
"corpus_records": 20643,
"corpus_sha256": "4c34a3fbeddbe050431a5e8881d3a2ea56adf6af3448912e700a6f5e919eb76d",
"datasets": [
"openbmb/VisRAG-Ret-Test-ArxivQA",
"openbmb/VisRAG-Ret-Test-ChartQA",
"openbmb/VisRAG-Ret-Test-InfoVQA",
"openbmb/VisRAG-Ret-Test-MP-DocVQA",
"openbmb/VisRAG-Ret-Test-PlotQA",
"openbmb/VisRAG-Ret-Test-SlideVQA"
],
"holders": [
"arxivqa",
"chartqa",
"infovqa",
"mp-docvqa",
"plotqa",
"slidevqa"
],
"queries": 480,
"queries_answerable": 240,
"queries_cross_holder_unanswerable": 240,
"queries_sha256": "1c74488375c7ec9f3182964d51a5e506273bfd5cf3d126725c3cc431a8333128",
"subsets": {
"arxivqa": {
"corpus_images": 8066,
"dataset": "openbmb/VisRAG-Ret-Test-ArxivQA",
"queries_annotated": 816,
"queries_dropped_missing_gold_image": 0,
"queries_usable": 816,
"revision": "00a31fc1e899b5f81d7760eca4907b1a1005865b"
},
"chartqa": {
"corpus_images": 500,
"dataset": "openbmb/VisRAG-Ret-Test-ChartQA",
"queries_annotated": 63,
"queries_dropped_missing_gold_image": 0,
"queries_usable": 63,
"revision": "31f5ceb5d60b02e065bff394cb582f5bbb01a9b6"
},
"infovqa": {
"corpus_images": 459,
"dataset": "openbmb/VisRAG-Ret-Test-InfoVQA",
"queries_annotated": 718,
"queries_dropped_missing_gold_image": 0,
"queries_usable": 718,
"revision": "04ee2db1d0a05304899453a3b03baef5e18991a3"
},
"mp-docvqa": {
"corpus_images": 741,
"dataset": "openbmb/VisRAG-Ret-Test-MP-DocVQA",
"queries_annotated": 591,
"queries_dropped_missing_gold_image": 0,
"queries_usable": 591,
"revision": "3ebd091c458cf04161f78cd7b12ea101f83e2529"
},
"plotqa": {
"corpus_images": 9593,
"dataset": "openbmb/VisRAG-Ret-Test-PlotQA",
"queries_annotated": 863,
"queries_dropped_missing_gold_image": 0,
"queries_usable": 863,
"revision": "ef953ef8ab6d78ac112dd4cde6acdb2c2692039a"
},
"slidevqa": {
"corpus_images": 1284,
"dataset": "openbmb/VisRAG-Ret-Test-SlideVQA",
"queries_annotated": 556,
"queries_dropped_missing_gold_image": 0,
"queries_usable": 556,
"revision": "6b5f81a3f6953095934c26c5d420dfd039184d69"
}
}
}