dataset.json 839 B

1234567891011
  1. {
  2. "dataset_version": "engineering-fixture-v1",
  3. "purpose": "local contract metric calculation only",
  4. "train_case_ids": ["train-001", "train-002"],
  5. "evaluation_cases": [
  6. {"case_id": "eval-001", "expected_evidence_id": "evidence-001", "retrieved_evidence_ids": ["evidence-001"], "citation_evidence_ids": ["evidence-001"], "authorized": true, "fresh": true},
  7. {"case_id": "eval-002", "expected_evidence_id": "evidence-002", "retrieved_evidence_ids": ["evidence-002", "evidence-003"], "citation_evidence_ids": ["evidence-002"], "authorized": true, "fresh": true},
  8. {"case_id": "eval-negative-001", "expected_evidence_id": "evidence-denied", "retrieved_evidence_ids": ["evidence-denied"], "citation_evidence_ids": ["evidence-denied"], "authorized": false, "fresh": false}
  9. ],
  10. "external_production_dataset": "TBD_EXTERNAL"
  11. }