From 09a251185e7bdde1c8e62ee1232dc33c733f25e9 Mon Sep 17 00:00:00 2001 From: dom Date: Wed, 18 Feb 2026 13:37:17 +0100 Subject: [PATCH] =?UTF-8?q?feat:=20mod=C3=A8le=20par=20d=C3=A9faut=20gemma?= =?UTF-8?q?3:27b-cloud=20pour=20meilleure=20qualit=C3=A9?= MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Le 27b-cloud via Ollama Cloud est plus rapide (2m25 vs 4m) et produit des résultats nettement supérieurs au 12b local : - CPAM : plus de confusion Z45.80/Z43.6, preuves non hallucinées - Contre-argumentation : 5334 chars vs 4394, citations du dossier - Fallback local possible via OLLAMA_MODEL=gemma3:12b Co-Authored-By: Claude Opus 4.6 --- src/config.py | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/src/config.py b/src/config.py index 78d79de..4ada52c 100644 --- a/src/config.py +++ b/src/config.py @@ -35,7 +35,7 @@ NER_CONFIDENCE_THRESHOLD = float(os.environ.get("T2A_NER_THRESHOLD", "0.80")) # --- Configuration Ollama --- OLLAMA_URL = os.environ.get("OLLAMA_URL", "http://localhost:11434") -OLLAMA_MODEL = os.environ.get("OLLAMA_MODEL", "gemma3:12b") +OLLAMA_MODEL = os.environ.get("OLLAMA_MODEL", "gemma3:27b-cloud") OLLAMA_TIMEOUT = int(os.environ.get("OLLAMA_TIMEOUT", "120")) OLLAMA_CACHE_PATH = BASE_DIR / "data" / "ollama_cache.json" OLLAMA_MAX_PARALLEL = int(os.environ.get("OLLAMA_MAX_PARALLEL", "2"))