{
 "about": "Every call type against gliner2 2.0.0 with the original weights (runtime/verify_parity.py): LiteRT 2.2.0 on the CPU, wfp16 encoder and the heads file.",
 "source": "fastino/GLiNER2.5-Decide@bbe10ff77ebb238777c17d3a8ac9260e30929057",
 "corpus": {
  "file": "runtime/parity_corpus.json",
  "requests": 296,
  "identical": 296,
  "byCallType": {
   "entities": {
    "calls": 40,
    "nativeError": 0,
    "refused": 0,
    "matched": 40,
    "items": 146,
    "minSpanF1": 1.0,
    "maxConfidenceDifference": 0.000214
   },
   "entities_single": {
    "calls": 8,
    "nativeError": 0,
    "refused": 0,
    "matched": 8,
    "items": 16,
    "minSpanF1": 1.0,
    "maxConfidenceDifference": 0.000236
   },
   "relations_work": {
    "calls": 24,
    "nativeError": 0,
    "refused": 0,
    "matched": 24,
    "items": 26,
    "minSpanF1": 1.0,
    "maxConfidenceDifference": 0.000351
   },
   "record_natural": {
    "calls": 16,
    "nativeError": 0,
    "refused": 0,
    "matched": 16,
    "items": 31,
    "minSpanF1": 1.0,
    "maxConfidenceDifference": 0.000303
   },
   "record_latent": {
    "calls": 24,
    "nativeError": 0,
    "refused": 0,
    "matched": 24,
    "items": 30,
    "minSpanF1": 1.0,
    "maxConfidenceDifference": 0.000489
   },
   "attributes": {
    "calls": 16,
    "nativeError": 0,
    "refused": 0,
    "matched": 16,
    "items": 26,
    "minSpanF1": 1.0,
    "maxConfidenceDifference": 0.00045
   },
   "classification_topic": {
    "calls": 40,
    "nativeError": 0,
    "refused": 0,
    "matched": 40,
    "items": 40,
    "minSpanF1": 1.0,
    "maxConfidenceDifference": 0.000204
   },
   "mixed": {
    "calls": 16,
    "nativeError": 0,
    "refused": 0,
    "matched": 16,
    "items": 55,
    "minSpanF1": 1.0,
    "maxConfidenceDifference": 0.000207
   },
   "relations_family": {
    "calls": 8,
    "nativeError": 0,
    "refused": 0,
    "matched": 8,
    "items": 8,
    "minSpanF1": 1.0,
    "maxConfidenceDifference": 0.000234
   },
   "classification_described": {
    "calls": 24,
    "nativeError": 0,
    "refused": 0,
    "matched": 24,
    "items": 24,
    "minSpanF1": 1.0,
    "maxConfidenceDifference": 9.3e-05
   },
   "record_anchorless": {
    "calls": 8,
    "nativeError": 0,
    "refused": 0,
    "matched": 8,
    "items": 12,
    "minSpanF1": 1.0,
    "maxConfidenceDifference": 0.000269
   },
   "structure_event": {
    "calls": 24,
    "nativeError": 0,
    "refused": 0,
    "matched": 24,
    "items": 38,
    "minSpanF1": 1.0,
    "maxConfidenceDifference": 0.001315
   },
   "classification_multi": {
    "calls": 16,
    "nativeError": 0,
    "refused": 0,
    "matched": 16,
    "items": 33,
    "minSpanF1": 1.0,
    "maxConfidenceDifference": 0.000334
   },
   "entities_described": {
    "calls": 16,
    "nativeError": 0,
    "refused": 0,
    "matched": 16,
    "items": 27,
    "minSpanF1": 1.0,
    "maxConfidenceDifference": 0.000182
   },
   "structure_order": {
    "calls": 8,
    "nativeError": 0,
    "refused": 0,
    "matched": 8,
    "items": 20,
    "minSpanF1": 1.0,
    "maxConfidenceDifference": 0.000579
   },
   "relations_event": {
    "calls": 8,
    "nativeError": 0,
    "refused": 0,
    "matched": 8,
    "items": 3,
    "minSpanF1": 1.0,
    "maxConfidenceDifference": 0.000118
   }
  }
 },
 "fixtures": {
  "about": "FluidInference's eleven verify-full-extraction fixtures",
  "identical": 11,
  "total": 11,
  "rows": [
   {
    "case": "entities",
    "structureMatch": true,
    "maxConfidenceDifference": 1e-06
   },
   {
    "case": "entities_multi",
    "structureMatch": true,
    "maxConfidenceDifference": 0.0
   },
   {
    "case": "relations",
    "structureMatch": true,
    "maxConfidenceDifference": 1e-06
   },
   {
    "case": "relations_two",
    "structureMatch": true,
    "maxConfidenceDifference": 2e-06
   },
   {
    "case": "record_natural",
    "structureMatch": true,
    "maxConfidenceDifference": 4e-06
   },
   {
    "case": "record_latent",
    "structureMatch": true,
    "maxConfidenceDifference": 4e-06
   },
   {
    "case": "record_anchorless",
    "structureMatch": true,
    "maxConfidenceDifference": 4e-06
   },
   {
    "case": "attributes",
    "structureMatch": true,
    "maxConfidenceDifference": 0.00014
   },
   {
    "case": "mixed_classification",
    "structureMatch": true,
    "maxConfidenceDifference": 5e-06
   },
   {
    "case": "classification_only",
    "structureMatch": true,
    "maxConfidenceDifference": 3e-05
   },
   {
    "case": "enum_choice",
    "structureMatch": true,
    "maxConfidenceDifference": 1e-06
   }
  ]
 },
 "spanFile": {
  "about": "gliner25_decide_span_wfp16.tflite and gliner25_decide_count_wfp16.tflite in place of the heads file (runtime/verify_parity.py --span --count), LiteRT 2.2.0 CPU",
  "corpus": {
   "requests": 296,
   "identical": 296,
   "byCallType": {
    "entities": {
     "calls": 40,
     "nativeError": 0,
     "refused": 0,
     "matched": 40,
     "items": 146,
     "minSpanF1": 1.0,
     "maxConfidenceDifference": 0.000214
    },
    "entities_single": {
     "calls": 8,
     "nativeError": 0,
     "refused": 0,
     "matched": 8,
     "items": 16,
     "minSpanF1": 1.0,
     "maxConfidenceDifference": 0.000236
    },
    "relations_work": {
     "calls": 24,
     "nativeError": 0,
     "refused": 0,
     "matched": 24,
     "items": 26,
     "minSpanF1": 1.0,
     "maxConfidenceDifference": 0.000351
    },
    "record_natural": {
     "calls": 16,
     "nativeError": 0,
     "refused": 0,
     "matched": 16,
     "items": 31,
     "minSpanF1": 1.0,
     "maxConfidenceDifference": 0.000303
    },
    "record_latent": {
     "calls": 24,
     "nativeError": 0,
     "refused": 0,
     "matched": 24,
     "items": 30,
     "minSpanF1": 1.0,
     "maxConfidenceDifference": 0.000489
    },
    "attributes": {
     "calls": 16,
     "nativeError": 0,
     "refused": 0,
     "matched": 16,
     "items": 26,
     "minSpanF1": 1.0,
     "maxConfidenceDifference": 0.00045
    },
    "classification_topic": {
     "calls": 40,
     "nativeError": 0,
     "refused": 0,
     "matched": 40,
     "items": 40,
     "minSpanF1": 1.0,
     "maxConfidenceDifference": 0.000204
    },
    "mixed": {
     "calls": 16,
     "nativeError": 0,
     "refused": 0,
     "matched": 16,
     "items": 55,
     "minSpanF1": 1.0,
     "maxConfidenceDifference": 0.000207
    },
    "relations_family": {
     "calls": 8,
     "nativeError": 0,
     "refused": 0,
     "matched": 8,
     "items": 8,
     "minSpanF1": 1.0,
     "maxConfidenceDifference": 0.000234
    },
    "classification_described": {
     "calls": 24,
     "nativeError": 0,
     "refused": 0,
     "matched": 24,
     "items": 24,
     "minSpanF1": 1.0,
     "maxConfidenceDifference": 9.3e-05
    },
    "record_anchorless": {
     "calls": 8,
     "nativeError": 0,
     "refused": 0,
     "matched": 8,
     "items": 12,
     "minSpanF1": 1.0,
     "maxConfidenceDifference": 0.000269
    },
    "structure_event": {
     "calls": 24,
     "nativeError": 0,
     "refused": 0,
     "matched": 24,
     "items": 38,
     "minSpanF1": 1.0,
     "maxConfidenceDifference": 0.001315
    },
    "classification_multi": {
     "calls": 16,
     "nativeError": 0,
     "refused": 0,
     "matched": 16,
     "items": 33,
     "minSpanF1": 1.0,
     "maxConfidenceDifference": 0.000334
    },
    "entities_described": {
     "calls": 16,
     "nativeError": 0,
     "refused": 0,
     "matched": 16,
     "items": 27,
     "minSpanF1": 1.0,
     "maxConfidenceDifference": 0.000182
    },
    "structure_order": {
     "calls": 8,
     "nativeError": 0,
     "refused": 0,
     "matched": 8,
     "items": 20,
     "minSpanF1": 1.0,
     "maxConfidenceDifference": 0.000579
    },
    "relations_event": {
     "calls": 8,
     "nativeError": 0,
     "refused": 0,
     "matched": 8,
     "items": 3,
     "minSpanF1": 1.0,
     "maxConfidenceDifference": 0.000118
    }
   }
  },
  "fixtures": {
   "identical": 11,
   "total": 11
  }
 }
}