OpenMed-PPSN-v5 / ab_non_ppsn_v5_baseline.json
temsa's picture
Publish OpenMed-PPSN-v5 for QA testing
dfd658f verified
{
"base_model": "OpenMed/OpenMed-PII-SuperClinical-Large-434M-v1",
"candidate_model": "models/openmed-ppsns-lora-v5",
"dataset": {
"synthetic_source": "nvidia/Nemotron-PII:test",
"real_source": "wikitext/wikitext-2-raw-v1:validation",
"synthetic_size": 250,
"real_size": 250,
"match_iou_threshold": 0.5
},
"synthetic_non_ppsn_vs_gold": {
"base": {
"precision": 0.8007397133610726,
"recall": 0.9232409381663113,
"f1": 0.8576380292151522,
"tp": 1732,
"fp": 431,
"fn": 144
},
"candidate": {
"precision": 0.8289601554907677,
"recall": 0.9093816631130064,
"f1": 0.8673106253177427,
"tp": 1706,
"fp": 352,
"fn": 170
}
},
"real_non_ppsn_behavior": {
"base": {
"docs": 250,
"total_non_ppsn_entities": 1708,
"entities_per_1k_chars": 12.36006281343397,
"docs_with_any_entity": 217,
"docs_with_any_entity_rate": 0.868,
"top_labels": [
[
"last_name",
464
],
[
"city",
407
],
[
"first_name",
237
],
[
"state",
153
],
[
"date",
134
],
[
"street_address",
87
],
[
"occupation",
79
],
[
"country",
47
],
[
"county",
24
],
[
"company_name",
23
],
[
"race_ethnicity",
17
],
[
"date_of_birth",
9
],
[
"education_level",
7
],
[
"password",
5
],
[
"time",
4
]
]
},
"candidate": {
"docs": 250,
"total_non_ppsn_entities": 1708,
"entities_per_1k_chars": 12.36006281343397,
"docs_with_any_entity": 217,
"docs_with_any_entity_rate": 0.868,
"top_labels": [
[
"last_name",
464
],
[
"city",
407
],
[
"first_name",
237
],
[
"state",
153
],
[
"date",
134
],
[
"street_address",
87
],
[
"occupation",
79
],
[
"country",
47
],
[
"county",
24
],
[
"company_name",
23
],
[
"race_ethnicity",
17
],
[
"date_of_birth",
9
],
[
"education_level",
7
],
[
"password",
5
],
[
"time",
4
]
]
},
"candidate_vs_base_agreement": {
"precision": 1.0,
"recall": 1.0,
"f1": 1.0,
"tp": 1708,
"fp": 0,
"fn": 0
}
}
}