Dom/t2a_v2
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

chore: add .gitignore

Browse Source
This commit is contained in:
dom
2026-03-05 00:37:41 +01:00
parent 542797a124
commit 2578afb6ff
1716 changed files with 1905609 additions and 18 deletions
Download Patch File Download Diff File Expand all files Collapse all files

32
output/benchmarks/20260220_005803_parallel_test/config.json Normal file
View File

@@ -0,0 +1,32 @@
{
"models": {
"coding": "gemma3:27b-cloud",
"cpam": "gemma3:27b-cloud",
"validation": "deepseek-v3.2:cloud",
"qc": "gemma3:12b"
},
"ollama_model": "gemma3:27b-cloud",
"ollama_url": "http://localhost:11434",
"timestamp": "2026-02-20T00:58:03.899793",
"run_id": "20260220_005803_parallel_test",
"dossiers": [
"116_23065570",
"132_23080179",
"134_23050890",
"143_23096917",
"145_23057452",
"153_23102610",
"170_23077016",
"176_23124187",
"17_23100690",
"183_23087212"
],
"args": {
"n": 10,
"gold_standard": true,
"clean": true,
"no_reprocess": false,
"seed": 42,
"label": "parallel_test"
}
}

839
output/benchmarks/20260220_005803_parallel_test/metrics.json Normal file
View File

@@ -0,0 +1,839 @@
{
"aggregate": {
"n_total": 10,
"n_success": 10,
"n_failed": 0,
"dp": {
"has_code_rate": 0.7,
"valid_code_rate": 0.7,
"confidence": {
"high": 5,
"medium": 2,
"low": 0
},
"downgraded": 0
},
"das": {
"total": 147,
"mean_per_dossier": 14.7,
"with_code": 147,
"valid": 147,
"validity_rate": 1.0,
"confidence": {
"high": 71,
"medium": 63,
"low": 13
},
"confidence_high_rate": 0.483,
"downgraded": 4,
"downgrade_rate": 0.027
},
"veto": {
"verdicts": {
"NEED_INFO": 8,
"PASS": 2
},
"hard_total": 0,
"dossiers_with_hard": 0
},
"ghm": {
"estimated_rate": 1.0
},
"cpam": {
"controls_total": 10,
"with_response": 10
},
"timing": {
"mean_s": 239.6,
"median_s": 222.4,
"p90_s": 317.7,
"total_s": 2396.4
},
"invalid_codes": [],
"invalid_codes_count": 0
},
"per_dossier": [
{
"dossier_id": "116_23065570",
"processing_time_s": 222.4,
"success": true,
"dp": {
"texte": "Insuffisance rénale aig",
"code_suggestion": "N17.9",
"code_final": "N17.9",
"confidence": "high",
"has_code": true,
"valid_code": true,
"downgraded": false
},
"das": {
"total": 9,
"with_code": 9,
"no_code": 0,
"valid": 9,
"validity_rate": 1.0,
"confidence": {
"high": 5,
"medium": 4,
"low": 0
},
"downgraded": 0,
"downgrade_rate": 0.0,
"codes_uniques": [
"K59.0",
"N13.9",
"N19",
"N21.0",
"N30.0",
"N40.0",
"N42.0",
"T83.5",
"Z01.4"
]
},
"metrics": {
"das_active": 9,
"das_removed": 0,
"das_ruled_out": 0
},
"veto": {
"verdict": "NEED_INFO",
"score": 78,
"issues_count": 5,
"hard_count": 0,
"top_issues": [
"VETO-03",
"VETO-03",
"VETO-03",
"VETO-03",
"VETO-03"
]
},
"ghm": {
"estimated": true,
"cmd": "11",
"severity": null,
"ghm": null
},
"cpam": {
"controls_count": 1,
"has_response": true,
"sources_count": 6
},
"biologie": {
"tests_count": 0,
"anomalies": 0
},
"invalid_codes": []
},
{
"dossier_id": "132_23080179",
"processing_time_s": 317.7,
"success": true,
"dp": {
"texte": "Adénopathie",
"code_suggestion": "R59.0",
"code_final": "R59.0",
"confidence": "medium",
"has_code": true,
"valid_code": true,
"downgraded": false
},
"das": {
"total": 22,
"with_code": 22,
"no_code": 0,
"valid": 22,
"validity_rate": 1.0,
"confidence": {
"high": 12,
"medium": 10,
"low": 0
},
"downgraded": 1,
"downgrade_rate": 0.045,
"codes_uniques": [
"C83.3",
"D64.9",
"D69.6",
"E66.84",
"E78.5",
"E86",
"E87.6",
"G44.8",
"I10",
"J90",
"M48.5",
"M54.2",
"M54.3",
"M54.8",
"M89.5",
"R06.0",
"R11",
"R13",
"R93.1",
"R93.8",
"T98.1",
"Z51.1"
]
},
"metrics": {
"das_active": 22,
"das_removed": 0,
"das_ruled_out": 0
},
"veto": {
"verdict": "NEED_INFO",
"score": 69,
"issues_count": 8,
"hard_count": 0,
"top_issues": [
"VETO-03",
"VETO-03",
"VETO-03",
"VETO-03",
"VETO-03"
]
},
"ghm": {
"estimated": true,
"cmd": "23",
"severity": null,
"ghm": null
},
"cpam": {
"controls_count": 1,
"has_response": true,
"sources_count": 6
},
"biologie": {
"tests_count": 12,
"anomalies": 7
},
"invalid_codes": []
},
{
"dossier_id": "134_23050890",
"processing_time_s": 249.7,
"success": true,
"dp": {
"texte": "Epispadia",
"code_suggestion": "Q64.0",
"code_final": "Q64.0",
"confidence": "high",
"has_code": true,
"valid_code": true,
"downgraded": false
},
"das": {
"total": 13,
"with_code": 13,
"no_code": 0,
"valid": 13,
"validity_rate": 1.0,
"confidence": {
"high": 5,
"medium": 6,
"low": 2
},
"downgraded": 0,
"downgrade_rate": 0.0,
"codes_uniques": [
"D50",
"E66.05",
"E66.9",
"F32.9",
"I95.1",
"J18.9",
"M16.9",
"M54.2",
"Q05.9",
"Q65.0",
"Q65.9",
"S73.1",
"Z04.802"
]
},
"metrics": {
"das_active": 13,
"das_removed": 0,
"das_ruled_out": 0
},
"veto": {
"verdict": "NEED_INFO",
"score": 75,
"issues_count": 6,
"hard_count": 0,
"top_issues": [
"VETO-03",
"VETO-03",
"VETO-03",
"VETO-03",
"VETO-03"
]
},
"ghm": {
"estimated": true,
"cmd": "15",
"severity": null,
"ghm": null
},
"cpam": {
"controls_count": 1,
"has_response": true,
"sources_count": 6
},
"biologie": {
"tests_count": 4,
"anomalies": 2
},
"invalid_codes": []
},
{
"dossier_id": "143_23096917",
"processing_time_s": 193.8,
"success": true,
"dp": {
"texte": "Appendicite aigue",
"code_suggestion": "K35.8",
"code_final": "K35.8",
"confidence": "medium",
"has_code": true,
"valid_code": true,
"downgraded": false
},
"das": {
"total": 14,
"with_code": 14,
"no_code": 0,
"valid": 14,
"validity_rate": 1.0,
"confidence": {
"high": 10,
"medium": 4,
"low": 0
},
"downgraded": 0,
"downgrade_rate": 0.0,
"codes_uniques": [
"B96.2",
"C69.4",
"D69.6",
"D72.8",
"E87.6",
"F06.4",
"G44.8",
"K65.9",
"K74.0",
"K74.3",
"L27.0",
"N17.9",
"R18",
"T80.8"
]
},
"metrics": {
"das_active": 13,
"das_removed": 0,
"das_ruled_out": 0
},
"veto": {
"verdict": "NEED_INFO",
"score": 78,
"issues_count": 5,
"hard_count": 0,
"top_issues": [
"VETO-03",
"VETO-03",
"VETO-03",
"VETO-03",
"VETO-15"
]
},
"ghm": {
"estimated": true,
"cmd": "06",
"severity": null,
"ghm": null
},
"cpam": {
"controls_count": 1,
"has_response": true,
"sources_count": 9
},
"biologie": {
"tests_count": 13,
"anomalies": 6
},
"invalid_codes": []
},
{
"dossier_id": "145_23057452",
"processing_time_s": 200.8,
"success": true,
"dp": {
"texte": "Examen général",
"code_suggestion": "Z00.8",
"code_final": "Z00.8",
"confidence": "high",
"has_code": true,
"valid_code": true,
"downgraded": false
},
"das": {
"total": 11,
"with_code": 11,
"no_code": 0,
"valid": 11,
"validity_rate": 1.0,
"confidence": {
"high": 4,
"medium": 6,
"low": 1
},
"downgraded": 0,
"downgrade_rate": 0.0,
"codes_uniques": [
"E11.8",
"F43.28",
"G44.8",
"G91.9",
"H40.9",
"I10",
"M10.9",
"M54.2",
"N18.3",
"N19",
"O86.0"
]
},
"metrics": {
"das_active": 11,
"das_removed": 0,
"das_ruled_out": 0
},
"veto": {
"verdict": "NEED_INFO",
"score": 75,
"issues_count": 6,
"hard_count": 0,
"top_issues": [
"VETO-03",
"VETO-03",
"VETO-03",
"VETO-03",
"VETO-09"
]
},
"ghm": {
"estimated": true,
"cmd": "23",
"severity": null,
"ghm": null
},
"cpam": {
"controls_count": 1,
"has_response": true,
"sources_count": 6
},
"biologie": {
"tests_count": 6,
"anomalies": 1
},
"invalid_codes": []
},
{
"dossier_id": "153_23102610",
"processing_time_s": 222.5,
"success": true,
"dp": {
"texte": "",
"code_suggestion": "",
"code_final": "",
"confidence": "",
"has_code": false,
"valid_code": false,
"downgraded": false
},
"das": {
"total": 15,
"with_code": 15,
"no_code": 0,
"valid": 15,
"validity_rate": 1.0,
"confidence": {
"high": 6,
"medium": 7,
"low": 2
},
"downgraded": 0,
"downgrade_rate": 0.0,
"codes_uniques": [
"C49",
"E11.9",
"E23.0",
"E78.5",
"F06.4",
"H36.0",
"K42.9",
"K71",
"L83",
"N19",
"N39.0",
"R59.9",
"R63.4",
"T83.5",
"T88.7"
]
},
"metrics": {
"das_active": 15,
"das_removed": 0,
"das_ruled_out": 0
},
"veto": {
"verdict": "NEED_INFO",
"score": 61,
"issues_count": 6,
"hard_count": 0,
"top_issues": [
"VETO-03",
"VETO-03",
"VETO-03",
"VETO-03",
"VETO-03"
]
},
"ghm": {
"estimated": true,
"cmd": null,
"severity": null,
"ghm": null
},
"cpam": {
"controls_count": 1,
"has_response": true,
"sources_count": 11
},
"biologie": {
"tests_count": 11,
"anomalies": 2
},
"invalid_codes": []
},
{
"dossier_id": "170_23077016",
"processing_time_s": 200.3,
"success": true,
"dp": {
"texte": "",
"code_suggestion": "",
"code_final": "",
"confidence": "",
"has_code": false,
"valid_code": false,
"downgraded": false
},
"das": {
"total": 11,
"with_code": 11,
"no_code": 0,
"valid": 11,
"validity_rate": 1.0,
"confidence": {
"high": 4,
"medium": 6,
"low": 1
},
"downgraded": 1,
"downgrade_rate": 0.091,
"codes_uniques": [
"D64.9",
"D72.8",
"I97.8",
"J18.9",
"K59.0",
"M05.9",
"N19",
"R23.1",
"R50.9",
"R52.2",
"Z00.0"
]
},
"metrics": {
"das_active": 11,
"das_removed": 0,
"das_ruled_out": 0
},
"veto": {
"verdict": "PASS",
"score": 94,
"issues_count": 2,
"hard_count": 0,
"top_issues": [
"VETO-03",
"VETO-03"
]
},
"ghm": {
"estimated": true,
"cmd": null,
"severity": null,
"ghm": null
},
"cpam": {
"controls_count": 1,
"has_response": true,
"sources_count": 6
},
"biologie": {
"tests_count": 6,
"anomalies": 2
},
"invalid_codes": []
},
{
"dossier_id": "176_23124187",
"processing_time_s": 293.9,
"success": true,
"dp": {
"texte": "Hta",
"code_suggestion": "I10",
"code_final": "I10",
"confidence": "high",
"has_code": true,
"valid_code": true,
"downgraded": false
},
"das": {
"total": 23,
"with_code": 23,
"no_code": 0,
"valid": 23,
"validity_rate": 1.0,
"confidence": {
"high": 13,
"medium": 8,
"low": 2
},
"downgraded": 1,
"downgrade_rate": 0.043,
"codes_uniques": [
"A41.9",
"C31.1",
"C34.9",
"D64.9",
"D69.6",
"E78.5",
"E83.5",
"E87.6",
"G00",
"I11.0",
"J18.9",
"J47",
"J90",
"K74.0",
"M06.4",
"N19",
"R05",
"R45.4",
"R59.9",
"R68.3",
"R93.8",
"Z72.0",
"Z99.1"
]
},
"metrics": {
"das_active": 22,
"das_removed": 0,
"das_ruled_out": 0
},
"veto": {
"verdict": "NEED_INFO",
"score": 59,
"issues_count": 9,
"hard_count": 0,
"top_issues": [
"VETO-03",
"VETO-03",
"VETO-03",
"VETO-03",
"VETO-03"
]
},
"ghm": {
"estimated": true,
"cmd": "05",
"severity": null,
"ghm": null
},
"cpam": {
"controls_count": 1,
"has_response": true,
"sources_count": 6
},
"biologie": {
"tests_count": 16,
"anomalies": 9
},
"invalid_codes": []
},
{
"dossier_id": "17_23100690",
"processing_time_s": 298.8,
"success": true,
"dp": {
"texte": "Pancréatite aiguë",
"code_suggestion": "K85.9",
"code_final": "K85.9",
"confidence": "high",
"has_code": true,
"valid_code": true,
"downgraded": false
},
"das": {
"total": 20,
"with_code": 20,
"no_code": 0,
"valid": 20,
"validity_rate": 1.0,
"confidence": {
"high": 7,
"medium": 9,
"low": 4
},
"downgraded": 1,
"downgrade_rate": 0.05,
"codes_uniques": [
"D64.9",
"E66.04",
"E78.5",
"E87.8",
"F10.1",
"G44.8",
"I10",
"K66.0",
"K71.8",
"K80.0",
"K80.1",
"K80.2",
"K83.0",
"N19",
"N85.7",
"R19",
"R52.10",
"R73.9",
"R82.4",
"Z00.0"
]
},
"metrics": {
"das_active": 20,
"das_removed": 0,
"das_ruled_out": 0
},
"veto": {
"verdict": "NEED_INFO",
"score": 68,
"issues_count": 6,
"hard_count": 0,
"top_issues": [
"VETO-03",
"VETO-03",
"VETO-03",
"VETO-03",
"VETO-03"
]
},
"ghm": {
"estimated": true,
"cmd": "07",
"severity": null,
"ghm": null
},
"cpam": {
"controls_count": 1,
"has_response": true,
"sources_count": 6
},
"biologie": {
"tests_count": 19,
"anomalies": 16
},
"invalid_codes": []
},
{
"dossier_id": "183_23087212",
"processing_time_s": 196.5,
"success": true,
"dp": {
"texte": "",
"code_suggestion": "",
"code_final": "",
"confidence": "",
"has_code": false,
"valid_code": false,
"downgraded": false
},
"das": {
"total": 9,
"with_code": 9,
"no_code": 0,
"valid": 9,
"validity_rate": 1.0,
"confidence": {
"high": 5,
"medium": 3,
"low": 1
},
"downgraded": 0,
"downgrade_rate": 0.0,
"codes_uniques": [
"E66.04",
"I10",
"N10",
"N19",
"N20.0",
"R35",
"T83.1",
"T83.5",
"Z43.3"
]
},
"metrics": {
"das_active": 9,
"das_removed": 0,
"das_ruled_out": 0
},
"veto": {
"verdict": "PASS",
"score": 91,
"issues_count": 3,
"hard_count": 0,
"top_issues": [
"VETO-03",
"VETO-03",
"VETO-03"
]
},
"ghm": {
"estimated": true,
"cmd": null,
"severity": null,
"ghm": null
},
"cpam": {
"controls_count": 1,
"has_response": true,
"sources_count": 0
},
"biologie": {
"tests_count": 6,
"anomalies": 0
},
"invalid_codes": []
}
]
}

72
output/benchmarks/20260220_005803_parallel_test/report.txt Normal file
View File

@@ -0,0 +1,72 @@
==================================================================
BENCHMARK QUALITÉ T2A — 20260220_005803_parallel_test
==================================================================
Date : 2026-02-20T00:58:03.899793
Modèles : coding=gemma3:27b-cloud cpam=gemma3:27b-cloud
validation=deepseek-v3.2:cloud qc=gemma3:12b
Dossiers : 10/10 traités (0 échecs)
Durée : 2396s (239.6s/dossier)
------------------------------------------------------------------
DIAGNOSTIC PRINCIPAL (DP)
Code obtenu : ██████████████░░░░░░ 70.0%
Code CIM-10 valide : ██████████████░░░░░░ 70.0%
Confiance high : 5/10 medium: 2 low: 0
Downgrades : 0
DIAGNOSTICS ASSOCIÉS (DAS)
Total : 147 (moy 14.7/dossier)
Avec code : 147/147
Codes valides : ████████████████████ 100.0%
Confiance : high=71 medium=63 low=13
Confiance high : █████████░░░░░░░░░░░ 48.3%
Downgrades : 4 (2.7%)
VETOS / QUALITÉ
NEED_INFO : 8
PASS : 2
Issues HARD : 0 (dans 0 dossiers)
GHM
Estimé : ████████████████████ 100.0%
CPAM
Contrôles : 10
Avec réponse : 10
TEMPS DE TRAITEMENT
Moyen : 239.6s
Médian : 222.4s
P90 : 317.7s
Total : 2396s
------------------------------------------------------------------
DÉTAIL PAR DOSSIER
------------------------------------------------------------------
Dossier DP DAS Valid% Veto Temps
------------------------- ------ ---- ------- ---------- ------
116_23065570 N17.9✓ 9 100% NEED_INFO 222s
132_23080179 R59.0✓ 22 100% NEED_INFO 318s
134_23050890 Q64.0✓ 13 100% NEED_INFO 250s
143_23096917 K35.8✓ 14 100% NEED_INFO 194s
145_23057452 Z00.8✓ 11 100% NEED_INFO 201s
153_23102610 -✗ 15 100% NEED_INFO 222s
170_23077016 -✗ 11 100% PASS 200s
176_23124187 I10✓ 23 100% NEED_INFO 294s
17_23100690 K85.9✓ 20 100% NEED_INFO 299s
183_23087212 -✗ 9 100% PASS 196s
==================================================================
==================================================================
COMPARAISON avec 20260220_005758_baseline_pre_parallel
==================================================================
Métrique Baseline Actuel Delta
------------------------ ---------- ---------- ----------
DP code valide 70.0% 70.0% +0.0%
DAS validité 100.0% 100.0% +0.0%
DAS confiance high 53.7% 48.3% -5.4%
DAS downgrade 2.7% 2.7% +0.0%
GHM estimé 100.0% 100.0% +0.0%
DAS moy/dossier 14.9 14.7 -0.2
Temps moyen (s) 0.0 239.6 +239.6
==================================================================