version.json

{
  "version": 3,
  "datasetTokens": 191488840,
  "datasetPages": 134994,
  "params": 122552666,
  "trainedAt": "2026-10-06T13:44:05+00:00",
  "weightsUrl": "https://huggingface.co/Crawlnet/queen-v3",
  "name": "queen-v3",
  "paramsNonEmbedding": 59638060,
  "tokensTrained": 625999872,
  "epochs": 3.27,
  "valBpb": 0.8173534432759958,
  "sftMethod": "synthetic+openbook:deepseek-flash",
  "datasetManifestSha256": "00a69452525b0533fdddde5ff0f345ac19d451cd5cc842b4711a0cfd833edea3",
  "costUsd": 0.6,
  "gpuHours": 1.771,
  "gpu": "1 GPU (our GPU server)",
  "nanochatCommit": "92d63d4e8bb4df75c3b71618f31ddde2378b2bcd",
  "toy": false,
  "tokenizer": "trained on this dataset only (crawler pages and record pages)",
  "license": "apache-2.0",
  "datasetSources": {
    "crawl": {
      "pages": 89367
    },
    "records": {
      "pages": 45627,
      "tokensApprox": 10950432,
      "what": "public market, funding and on-chain records (projects and their funding rounds, investors, token unlocks and token facts, DeFi protocols, chains and stablecoins, DEX and on-chain token statistics)"
    }
  }
}