From 95135a665b29d53544192f1c3e6c8dc03ba9b445 Mon Sep 17 00:00:00 2001 From: Bryan Van Deusen Date: Sun, 12 Apr 2026 15:45:22 -0400 Subject: [PATCH] fix(llm): switch default background model to gemma3:4b MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit qwen2.5:3b produced broken auto-summaries (misspellings, token repetition, hallucinated topics) — its synthesis ceiling is too low for free-form summarization. Gemma 3 4B is stronger on summarization at similar size and still fits comfortably alongside the main chat model in VRAM, so it preserves the KV-cache-separation strategy that keeps chat TTFT fast. Co-Authored-By: Claude Opus 4.6 --- src/fabledassistant/config.py | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/src/fabledassistant/config.py b/src/fabledassistant/config.py index ebb9802..7c447d0 100644 --- a/src/fabledassistant/config.py +++ b/src/fabledassistant/config.py @@ -27,7 +27,7 @@ class Config: # Lightweight model for background tasks (title generation, tag suggestions, # project summaries, RSS classification). Using a separate model keeps the # main model's KV cache intact between user messages, enabling prefix cache hits. - OLLAMA_BACKGROUND_MODEL: str = os.environ.get("OLLAMA_BACKGROUND_MODEL", "qwen2.5:3b") + OLLAMA_BACKGROUND_MODEL: str = os.environ.get("OLLAMA_BACKGROUND_MODEL", "gemma3:4b") # Ollama keep_alive — how long a model stays resident in VRAM after its last # request. Main model gets a longer window since it's used interactively; # the background model is called sporadically and doesn't need to camp VRAM.