{
  "schema_version": "2.0",
  "record_type": "dataset",
  "tier": "verified",
  "slug": "medqa-diag",
  "name": "R2MED Benchmark",
  "summary": {
    "en": "License: permissive (CC-BY-4.0); Mdexa hosting: link-only. first-hand data-scope statement(s) allow commercial use and redistribution",
    "zh": "许可证声明：宽松（CC-BY-4.0）；本站处理：只链接原出处。针对数据的声明允许商用和再分发。"
  },
  "systems": [
    "modern-medicine"
  ],
  "domains": [
    "clinical-medicine"
  ],
  "content_types": [
    "annotated-corpus"
  ],
  "content_description": {
    "en": "R2MED is a high-quality synthetic medical information retrieval dataset with 876 queries, three retrieval tasks, five medical scenarios and twelve body systems.",
    "zh": "R2MED是一个高质量的合成医学信息检索数据集，包含876条查询、三个检索任务、五个医学场景和十二个身体系统。"
  },
  "languages": [
    "en"
  ],
  "modalities": [
    "text"
  ],
  "size": {
    "raw": "10M<n<100M",
    "measures": []
  },
  "publisher": "R2MED",
  "year": {
    "value": 2025,
    "raw": "2025-05-21"
  },
  "license": {
    "legacy_status": "pending",
    "spdx": "CC-BY-4.0",
    "commercial_use": "yes",
    "derivatives": "yes",
    "redistribution": "yes",
    "rationale": {
      "en": "first-hand data-scope statement(s) allow commercial use and redistribution",
      "zh": "针对数据的声明允许商用和再分发"
    },
    "evidence": [
      {
        "location": "hf-card-metadata",
        "location_detail": {
          "en": "Statement found in snapshot s04 (readme)",
          "zh": "在快照 s04（readme）中找到的声明"
        },
        "excerpt": "license: cc-by-4.0",
        "spdx": "CC-BY-4.0",
        "url": "https://huggingface.co/datasets/R2MED/MedQA-Diag",
        "verified_on": "2026-11-12",
        "snapshot_id": "s04",
        "observed": "page",
        "captured_by": "pipeline"
      }
    ],
    "license_conditions": [
      "attribution"
    ]
  },
  "license_status": "permissive",
  "hosting": "link-only",
  "hosting_reason": {
    "en": "License permissive: first-hand data-scope statement(s) allow commercial use and redistribution. Link only.",
    "zh": "许可证宽松：针对数据的声明允许商用和再分发。只链接原出处。"
  },
  "signoff": null,
  "risk_flags": [],
  "provenance": {
    "source_description": null,
    "upstream": []
  },
  "known_issues": [],
  "preflight": {
    "status": "not-run"
  },
  "access": {
    "mdexa_download": null,
    "external_urls": [
      "https://huggingface.co/datasets/R2MED/MedQA-Diag"
    ],
    "notes": null
  },
  "citation": {
    "text": null,
    "references": [
      "arXiv:2505.14558"
    ]
  },
  "verification_level": "verified",
  "last_verified": "2026-11-14",
  "tags": [
    {
      "facet": "intended_use",
      "value": "evaluation",
      "basis": "inferred",
      "inferred_from": "card-content-description",
      "excerpt": "### R2MED: First Reasoning-Driven Medical Retrieval Benchmark",
      "snapshot_id": "s04"
    },
    {
      "facet": "intended_use",
      "value": "retrieval-kb",
      "basis": "inferred",
      "inferred_from": "snapshot-structure",
      "excerpt": "      - split: corpus\n        path: corpus.jsonl",
      "snapshot_id": "s04"
    },
    {
      "facet": "specialty",
      "value": "multi-specialty",
      "basis": "inferred",
      "inferred_from": "card-content-description",
      "excerpt": "It contains 876 queries with three retrieval tasks, five medical scenarios, and twelve body systems.",
      "snapshot_id": "s04"
    },
    {
      "facet": "data_format",
      "value": "jsonl",
      "basis": "inferred",
      "inferred_from": "snapshot-structure",
      "excerpt": "        path: qrels.jsonl",
      "snapshot_id": "s04"
    },
    {
      "facet": "benchmark_status",
      "value": "public-leaderboard",
      "basis": "author-stated",
      "excerpt": "You could check out the results at [Leaderboard](https://r2med.github.io/#leaderboard).",
      "snapshot_id": "s04"
    }
  ],
  "changelog": [
    {
      "date": "2026-11-14",
      "change": {
        "en": "Added via the automated pipeline and passed all checks (gates, verbatim quotes, rule engine, two independent extractions).",
        "zh": "经自动流水线收录并通过全部校验（门禁、逐字引用、判定引擎、两轮独立抽取）。"
      }
    }
  ],
  "access_recipe": {
    "verified_on": "2026-11-14",
    "platform": "huggingface",
    "repo_id": "R2MED/MedQA-Diag",
    "source": {
      "api_url": "https://huggingface.co/api/datasets/R2MED/MedQA-Diag",
      "fetched_on": "2026-11-14"
    },
    "access_state": "open",
    "files": [
      {
        "name": ".gitattributes",
        "url": "https://huggingface.co/datasets/R2MED/MedQA-Diag/resolve/78b585990279cc01a493f876c1b0cf09557fba57/.gitattributes",
        "size_bytes": 2510,
        "checksum": null,
        "format": "gitattributes"
      },
      {
        "name": "README.md",
        "url": "https://huggingface.co/datasets/R2MED/MedQA-Diag/resolve/78b585990279cc01a493f876c1b0cf09557fba57/README.md",
        "size_bytes": 3890,
        "checksum": null,
        "format": "md"
      },
      {
        "name": "corpus.jsonl",
        "url": "https://huggingface.co/datasets/R2MED/MedQA-Diag/resolve/78b585990279cc01a493f876c1b0cf09557fba57/corpus.jsonl",
        "size_bytes": 46866718,
        "checksum": {
          "algorithm": "sha256",
          "value": "2767063e36589bcda9ea87a898effd91529548838d645adccbc9e9c22bdd0f12"
        },
        "format": "jsonl"
      },
      {
        "name": "hyde/gpt4/query_with_hydoc.jsonl",
        "url": "https://huggingface.co/datasets/R2MED/MedQA-Diag/resolve/78b585990279cc01a493f876c1b0cf09557fba57/hyde/gpt4/query_with_hydoc.jsonl",
        "size_bytes": 334332,
        "checksum": null,
        "format": "jsonl"
      },
      {
        "name": "hyde/huatuo-o1-70b/query_with_hydoc.jsonl",
        "url": "https://huggingface.co/datasets/R2MED/MedQA-Diag/resolve/78b585990279cc01a493f876c1b0cf09557fba57/hyde/huatuo-o1-70b/query_with_hydoc.jsonl",
        "size_bytes": 441040,
        "checksum": null,
        "format": "jsonl"
      },
      {
        "name": "hyde/llama-70b/query_with_hydoc.jsonl",
        "url": "https://huggingface.co/datasets/R2MED/MedQA-Diag/resolve/78b585990279cc01a493f876c1b0cf09557fba57/hyde/llama-70b/query_with_hydoc.jsonl",
        "size_bytes": 283185,
        "checksum": null,
        "format": "jsonl"
      },
      {
        "name": "hyde/o3-mini/query_with_hydoc.jsonl",
        "url": "https://huggingface.co/datasets/R2MED/MedQA-Diag/resolve/78b585990279cc01a493f876c1b0cf09557fba57/hyde/o3-mini/query_with_hydoc.jsonl",
        "size_bytes": 253298,
        "checksum": null,
        "format": "jsonl"
      },
      {
        "name": "hyde/qwen-32b/query_with_hydoc.jsonl",
        "url": "https://huggingface.co/datasets/R2MED/MedQA-Diag/resolve/78b585990279cc01a493f876c1b0cf09557fba57/hyde/qwen-32b/query_with_hydoc.jsonl",
        "size_bytes": 349476,
        "checksum": null,
        "format": "jsonl"
      },
      {
        "name": "hyde/qwen-72b/query_with_hydoc.jsonl",
        "url": "https://huggingface.co/datasets/R2MED/MedQA-Diag/resolve/78b585990279cc01a493f876c1b0cf09557fba57/hyde/qwen-72b/query_with_hydoc.jsonl",
        "size_bytes": 377749,
        "checksum": null,
        "format": "jsonl"
      },
      {
        "name": "hyde/qwen-7b/query_with_hydoc.jsonl",
        "url": "https://huggingface.co/datasets/R2MED/MedQA-Diag/resolve/78b585990279cc01a493f876c1b0cf09557fba57/hyde/qwen-7b/query_with_hydoc.jsonl",
        "size_bytes": 332022,
        "checksum": null,
        "format": "jsonl"
      },
      {
        "name": "hyde/qwq-32b/query_with_hydoc.jsonl",
        "url": "https://huggingface.co/datasets/R2MED/MedQA-Diag/resolve/78b585990279cc01a493f876c1b0cf09557fba57/hyde/qwq-32b/query_with_hydoc.jsonl",
        "size_bytes": 1157540,
        "checksum": null,
        "format": "jsonl"
      },
      {
        "name": "hyde/r1-llama-70b/query_with_hydoc.jsonl",
        "url": "https://huggingface.co/datasets/R2MED/MedQA-Diag/resolve/78b585990279cc01a493f876c1b0cf09557fba57/hyde/r1-llama-70b/query_with_hydoc.jsonl",
        "size_bytes": 658342,
        "checksum": null,
        "format": "jsonl"
      },
      {
        "name": "hyde/r1-qwen-32b/query_with_hydoc.jsonl",
        "url": "https://huggingface.co/datasets/R2MED/MedQA-Diag/resolve/78b585990279cc01a493f876c1b0cf09557fba57/hyde/r1-qwen-32b/query_with_hydoc.jsonl",
        "size_bytes": 618656,
        "checksum": null,
        "format": "jsonl"
      },
      {
        "name": "lamer/gpt4/query_with_hydoc.jsonl",
        "url": "https://huggingface.co/datasets/R2MED/MedQA-Diag/resolve/78b585990279cc01a493f876c1b0cf09557fba57/lamer/gpt4/query_with_hydoc.jsonl",
        "size_bytes": 179132,
        "checksum": null,
        "format": "jsonl"
      },
      {
        "name": "lamer/qwen-72b/query_with_hydoc.jsonl",
        "url": "https://huggingface.co/datasets/R2MED/MedQA-Diag/resolve/78b585990279cc01a493f876c1b0cf09557fba57/lamer/qwen-72b/query_with_hydoc.jsonl",
        "size_bytes": 191186,
        "checksum": null,
        "format": "jsonl"
      },
      {
        "name": "lamer/qwen-7b/query_with_hydoc.jsonl",
        "url": "https://huggingface.co/datasets/R2MED/MedQA-Diag/resolve/78b585990279cc01a493f876c1b0cf09557fba57/lamer/qwen-7b/query_with_hydoc.jsonl",
        "size_bytes": 178606,
        "checksum": null,
        "format": "jsonl"
      },
      {
        "name": "qrels.jsonl",
        "url": "https://huggingface.co/datasets/R2MED/MedQA-Diag/resolve/78b585990279cc01a493f876c1b0cf09557fba57/qrels.jsonl",
        "size_bytes": 32662,
        "checksum": null,
        "format": "jsonl"
      },
      {
        "name": "query.jsonl",
        "url": "https://huggingface.co/datasets/R2MED/MedQA-Diag/resolve/78b585990279cc01a493f876c1b0cf09557fba57/query.jsonl",
        "size_bytes": 110103,
        "checksum": null,
        "format": "jsonl"
      },
      {
        "name": "query2doc/gpt4/query_with_hydoc.jsonl",
        "url": "https://huggingface.co/datasets/R2MED/MedQA-Diag/resolve/78b585990279cc01a493f876c1b0cf09557fba57/query2doc/gpt4/query_with_hydoc.jsonl",
        "size_bytes": 291876,
        "checksum": null,
        "format": "jsonl"
      },
      {
        "name": "query2doc/qwen-72b/query_with_hydoc.jsonl",
        "url": "https://huggingface.co/datasets/R2MED/MedQA-Diag/resolve/78b585990279cc01a493f876c1b0cf09557fba57/query2doc/qwen-72b/query_with_hydoc.jsonl",
        "size_bytes": 387174,
        "checksum": null,
        "format": "jsonl"
      },
      {
        "name": "query2doc/qwen-7b/query_with_hydoc.jsonl",
        "url": "https://huggingface.co/datasets/R2MED/MedQA-Diag/resolve/78b585990279cc01a493f876c1b0cf09557fba57/query2doc/qwen-7b/query_with_hydoc.jsonl",
        "size_bytes": 329462,
        "checksum": null,
        "format": "jsonl"
      },
      {
        "name": "search-o1/qwen3-32b/query_with_hydoc.jsonl",
        "url": "https://huggingface.co/datasets/R2MED/MedQA-Diag/resolve/78b585990279cc01a493f876c1b0cf09557fba57/search-o1/qwen3-32b/query_with_hydoc.jsonl",
        "size_bytes": 1169275,
        "checksum": null,
        "format": "jsonl"
      },
      {
        "name": "search-o1/qwq-32b/query_with_hydoc.jsonl",
        "url": "https://huggingface.co/datasets/R2MED/MedQA-Diag/resolve/78b585990279cc01a493f876c1b0cf09557fba57/search-o1/qwq-32b/query_with_hydoc.jsonl",
        "size_bytes": 1629267,
        "checksum": null,
        "format": "jsonl"
      },
      {
        "name": "search-r1/qwen-3b-ins/query_with_hydoc.jsonl",
        "url": "https://huggingface.co/datasets/R2MED/MedQA-Diag/resolve/78b585990279cc01a493f876c1b0cf09557fba57/search-r1/qwen-3b-ins/query_with_hydoc.jsonl",
        "size_bytes": 2183434,
        "checksum": null,
        "format": "jsonl"
      },
      {
        "name": "search-r1/qwen-7b-ins/query_with_hydoc.jsonl",
        "url": "https://huggingface.co/datasets/R2MED/MedQA-Diag/resolve/78b585990279cc01a493f876c1b0cf09557fba57/search-r1/qwen-7b-ins/query_with_hydoc.jsonl",
        "size_bytes": 1637739,
        "checksum": null,
        "format": "jsonl"
      }
    ],
    "files_truncated": false,
    "commands": [
      {
        "command": "hf download R2MED/MedQA-Diag --repo-type dataset",
        "evidence": {
          "excerpt": "You can use the hf download command from the terminal to directly download files from the Hub.",
          "doc": "hf-download",
          "url": "https://huggingface.co/docs/huggingface_hub/guides/download",
          "fetched_on": "2026-11-14"
        }
      }
    ],
    "registration": null,
    "version": {
      "commit": "78b585990279cc01a493f876c1b0cf09557fba57",
      "last_modified": "2025-05-21T18:11:34.000Z"
    },
    "license_note": {
      "license_status": "permissive",
      "text": "Use according to the terms of the official source; this recipe does not say the data can be used freely."
    },
    "gaps": []
  },
  "_links": {
    "text": "在本库继续查找。本条目是 109 条已处理并核实的记录之一（数据集、数据资源与开源模型）。查找其他条目：调用 `https://mcp.mdexa.com/v1/datasets?q=<关键词>`，或连接 MCP 服务 `https://mcp.mdexa.com/mcp`；每分钟 20 次以内无需密钥。每行返回不到 1 KB；空结果即\"数据库未收录\"。核对一份清单：一次批量核对（免费账号）即返回报告。说明：/agents.md。相似条目：PMC-Patients-ReCDS (https://mdexa.com/zh/datasets/pmc-patients-recds/)；PromptCBLUE (https://mdexa.com/zh/datasets/promptcblue/)；PubMedQA (https://mdexa.com/zh/datasets/pubmedqa/)。",
    "records": 109,
    "search": {
      "template": "https://mcp.mdexa.com/v1/datasets?q={words}",
      "example": "https://mcp.mdexa.com/v1/datasets?q=%E7%BB%93%E5%90%88%E4%BA%B2%E5%92%8C%E5%8A%9B"
    },
    "mcp": "https://mcp.mdexa.com/mcp",
    "anonymous_calls_per_minute": 20,
    "guide": "https://mdexa.com/zh/agents.md",
    "faq": "https://mdexa.com/zh/developers/faq/",
    "similar": [
      {
        "name": "PMC-Patients-ReCDS",
        "url": "https://mdexa.com/zh/datasets/pmc-patients-recds/",
        "json": "https://mdexa.com/zh/datasets/pmc-patients-recds.json"
      },
      {
        "name": "PromptCBLUE",
        "url": "https://mdexa.com/zh/datasets/promptcblue/",
        "json": "https://mdexa.com/zh/datasets/promptcblue.json"
      },
      {
        "name": "PubMedQA",
        "url": "https://mdexa.com/zh/datasets/pubmedqa/",
        "json": "https://mdexa.com/zh/datasets/pubmedqa.json"
      }
    ]
  }
}