{
  "title": "Historical Reasoning & Forestry Pilot",
  "category": "LLM training",
  "status": "PRELIMINARY",
  "version": "2026-10-23-preliminary-v1",
  "source_dataset": "PleIAs/common_corpus",
  "source_url": "https://huggingface.co/datasets/PleIAs/common_corpus",
  "source_revision": "307910e4c5d040d6f318e6edf2a2b97849155771",
  "downloaded_source_bytes": 429962586,
  "chunk_count": 37,
  "word_count": 8801,
  "upstream_source_fragment_count": 3,
  "source_work_count": 3,
  "book_count": 2,
  "journal_volume_count": 1,
  "topic_counts": {
    "reasoning_language": 30,
    "forestry_ecology": 7
  },
  "license": "Public-domain source text (United States assessment); CC0 for Krummholz additions to the extent rights can be granted",
  "description": "A small auditable sample of historical reasoning, language, forestry and botany prose for LLM data-pipeline and retrieval experiments.",
  "limitations": [
    "Not a representative training corpus or benchmark; no model training, evaluation or improvement claim.",
    "Residual OCR errors, historical inaccuracies and biased or unsafe historical practices may remain.",
    "Krummholz generated no new text or completions; upstream OCR/model corrections may remain.",
    "Private-contact patterns screened and all passages reviewed; not a guarantee of anonymization. Historical citations and geographic references remain.",
    "United States oriented public-domain assessment, not global clearance.",
    "Contains no fisheries or engineering excerpts."
  ],
  "website_bullet": "37 curated excerpts \u00b7 8,801 words \u00b7 3 source fragments \u00b7 PRELIMINARY",
  "jsonl": {
    "name": "krummholz-llm-training-preliminary.jsonl",
    "bytes": 92088,
    "sha256": "4e45ec4f38760328ac743db4ba43c4c10ad59018a1a92586d3affd3448a61433"
  },
  "zip": {
    "name": "krummholz-llm-training-preliminary-2026-10-23.zip",
    "bytes": 46791,
    "sha256": "52aae4b5a7a444ebea6f5c22078b4a32e8acaf16f3155a6d726008b2e01fffb5"
  },
  "files": [
    {
      "name": "audit.json",
      "bytes": 4380,
      "sha256": "138b112d5e002c6108e1d1a60f78a82e30faabfc99ed2e6b7c8447b410bde188"
    },
    {
      "name": "krummholz-llm-training-preliminary.jsonl",
      "bytes": 92088,
      "sha256": "4e45ec4f38760328ac743db4ba43c4c10ad59018a1a92586d3affd3448a61433"
    },
    {
      "name": "LICENSES.txt",
      "bytes": 2269,
      "sha256": "3d1bcd67bbde817f59c561276b27a4581ec0d7fd2d69270c29135aec291b6b6e"
    },
    {
      "name": "README.md",
      "bytes": 7006,
      "sha256": "d3866ea91c31e5507bff87401fe600d8fde9f5002a092a465493a6c6b66ba48d"
    },
    {
      "name": "reproduce/curate.py",
      "bytes": 15440,
      "sha256": "70ccaa2c5151b5ea5755ecce3aac0272c0ab73071a27e25aa3b9f6eedf8ae6c5"
    },
    {
      "name": "reproduce/download_source.py",
      "bytes": 1400,
      "sha256": "4e07d13f4b879f78cd77f346002e5361a0a0feb61f8f89d9c11d2b66db689851"
    },
    {
      "name": "reproduce/editorial-passage-selection.json",
      "bytes": 3656,
      "sha256": "c6800148edf9b8a165d3294a63fefb2432096289f83e180fcfad3543490d414c"
    },
    {
      "name": "reproduce/editorial-selection.json",
      "bytes": 1796,
      "sha256": "93f28796d9128a6377877ac9c2ff041d5fc20245238999f904832d4affd4db9a"
    },
    {
      "name": "reproduce/manual-exclusions.json",
      "bytes": 4583,
      "sha256": "ec96f9cbff1c13c2dac2befca0b925d3b8d7ec99b7c9a8812de3504b091ea525"
    },
    {
      "name": "reproduce/requirements.txt",
      "bytes": 17,
      "sha256": "0450cadcb3188b286b5a178c07ef4f1ddd70a9537be7d2d21a0e21ab16930218"
    },
    {
      "name": "reproduce/reviewed-sources.json",
      "bytes": 9054,
      "sha256": "437c149f8dc9e24c7400b20e43ccc16b1c79b1417172d26b9a3644d3e5b7cb7a"
    },
    {
      "name": "reproduce/source-download.json",
      "bytes": 318,
      "sha256": "1aa2e54dbefe913a4127905e15869548dead84ab12845f64c84009c74c53b8a3"
    },
    {
      "name": "schema.json",
      "bytes": 4025,
      "sha256": "e9c2f88800b115ca88a515c38e589c70f0ae5aab450ff49247921bce803aa968"
    },
    {
      "name": "sources.json",
      "bytes": 2796,
      "sha256": "e139387cb21848098904d6b299a3a3cb50a0056c91fd720bdaee658062a82879"
    }
  ]
}