You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Implement benchmarks/fixtures/provenance.py and gold.py. Gold fixture must include false-repair traps for IDs, targets, and free text so repair_fidelity and false_repair_rate metrics are grounded in known-correct ground truth.
Implement
benchmarks/fixtures/provenance.pyandgold.py. Gold fixture must include false-repair traps for IDs, targets, and free text so repair_fidelity and false_repair_rate metrics are grounded in known-correct ground truth.Implemented in #76.