Update Decision Engine and related components for WP-25a: Bump version to 1.2.0, enhance multi-stream retrieval with pre-synthesis compression, and integrate Mixture of Experts (MoE) profile support. Refactor chat interface to utilize new compression logic and llm_profiles for improved synthesis. Maintain compatibility with existing methods and ensure robust error handling across services.

This commit is contained in:
Lars committed 2026-01-02 07:04:43 +01:00
1 parent 3d2f3d12d9
commit d0eae8e43c
5 files changed
+299 -200

No files matched your search

+31
View File
@@ -0,0 +1,31 @@
# config/llm_profiles.yaml
# VERSION: 1.0.0 (WP-25a: Centralized MoE Profiles)
# STATUS: Active
# DESCRIPTION: Zentrale Definition der LLM-Experten-Profile für MindNet.
profiles:
# Der "Dampfhammer": Schnell und günstig für Zusammenfassungen
compression_fast:
provider: "openrouter"
model: "google/gemini-flash-1.5"
temperature: 0.1
# Der "Ingenieur": Tiefes Verständnis für Code und Logik
tech_expert:
provider: "openrouter"
model: "anthropic/claude-3-sonnet"
temperature: 0.3
# Der "Wächter": Lokal für sensible Identitäts-Daten
identity_safe:
provider: "ollama"
model: "llama3.1:8b"
temperature: 0.2
# Der "Architekt": Hochwertige Synthese und strategische Abwägung
synthesis_pro:
provider: "gemini"
model: "gemini-1.5-pro"
temperature: 0.7