{
  "schema_version": 1,
  "storage": {
    "qwen_model": "Qwen3.8-27B",
    "qwen_format": "Full BF16 weights",
    "qwen_weight_bytes": 55562855904,
    "qwen_revision": "1d4bf0f2ff6012fd82039f2fa52739d0dd7c60c0",
    "qwen_source": "https://huggingface.co/Qwen/Qwen3.8-27B/blob/1d4bf0f2ff6012fd82039f2fa52739d0dd7c60c0/model.safetensors.index.json",
    "larry_format": "Text-only Q1_0 GGUF",
    "larry_file_bytes": 3803453792,
    "larry_file_sha256": "df6679377b6007cf52fcd5685a4f1dc178e37766c1d131473cec6e92553356c4",
    "ratio": 14.608526603075397,
    "display_ratio": "14.6x",
    "circle_scaling": "area proportional to bytes"
  },
  "accuracy": {
    "benchmark": "IFEval",
    "metric": "mean_prompt_level_strict",
    "definition": "A prompt passes only if every instruction is followed.",
    "snapshot_sha256": "36ec91ffa20130e25937ae576369d91fa1cf9d9dc04f1c66c5cf38ebde5f876d",
    "larry": {
      "score": 0.7778,
      "prompts": 477,
      "source_report_sha256": "38db70ac8fce0e19b1f0f1e0558229f823875cb54e4fc9cba730bfd9b6810f5b",
      "all_metrics": [
        {
          "name": "mean_prompt_level_strict",
          "score": 0.7778,
          "num": 477
        },
        {
          "name": "mean_inst_level_strict",
          "score": 0.8484,
          "num": 477
        },
        {
          "name": "mean_prompt_level_loose",
          "score": 0.8176,
          "num": 477
        },
        {
          "name": "mean_inst_level_loose",
          "score": 0.8753,
          "num": 477
        }
      ]
    },
    "qwen": {
      "score": 0.847,
      "prompts": 477,
      "source_report_sha256": "645d67d76669e1f12de453de45f163e3c4c8159efdade82a7c5106100bdd28f1",
      "all_metrics": [
        {
          "name": "mean_prompt_level_strict",
          "score": 0.847,
          "num": 477
        },
        {
          "name": "mean_inst_level_strict",
          "score": 0.9022,
          "num": 477
        },
        {
          "name": "mean_prompt_level_loose",
          "score": 0.8868,
          "num": 477
        },
        {
          "name": "mean_inst_level_loose",
          "score": 0.9284,
          "num": 477
        }
      ]
    },
    "relative_accuracy": 0.9182998819362457,
    "display_relative_accuracy": "92%",
    "evaluation_period": "August 2026 archived runs",
    "scope": "A single instruction-following benchmark, not an overall capability estimate. Archived benchmark runs use different inference settings from the live chat."
  }
}