Lars/mindnet
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases 21 Wiki Activity

anpassung an variable Kontextgrenzen für Ollama
All checks were successful
Deploy mindnet to llm-node / deploy (push) Successful in 5s
Details

Browse Source
This commit is contained in:
Lars 2025-12-26 11:15:13 +01:00
parent 6ac1f318d0
commit f686ecf947
2 changed files with 9 additions and 4 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

2
app/config.py
View File

@ -34,6 +34,8 @@ class Settings:
# --- WP-20 Hybrid LLM Provider --- # --- WP-20 Hybrid LLM Provider ---
# Erlaubt: "ollama" | "gemini" | "openrouter" # Erlaubt: "ollama" | "gemini" | "openrouter"
MINDNET_LLM_PROVIDER: str = os.getenv("MINDNET_LLM_PROVIDER", "openrouter").lower() MINDNET_LLM_PROVIDER: str = os.getenv("MINDNET_LLM_PROVIDER", "openrouter").lower()
# Standardwert 10000, falls nichts in der .env steht
MAX_OLLAMA_CHARS: int = int(os.getenv("MAX_OLLAMA_CHARS", 10000))
# Google AI Studio (2025er Lite-Modell für höhere Kapazität) # Google AI Studio (2025er Lite-Modell für höhere Kapazität)
GOOGLE_API_KEY: str | None = os.getenv("GOOGLE_API_KEY") GOOGLE_API_KEY: str | None = os.getenv("GOOGLE_API_KEY")

11
app/routers/chat.py
View File

@ -310,10 +310,13 @@ async def chat_endpoint(
# --- STABILITY FIX: OLLAMA CONTEXT THROTTLE --- # --- STABILITY FIX: OLLAMA CONTEXT THROTTLE ---
# Begrenzt den Text, um den "decode: cannot decode batches" Fehler zu vermeiden. # Begrenzt den Text, um den "decode: cannot decode batches" Fehler zu vermeiden.
MAX_OLLAMA_CHARS = 10000 # MAX_OLLAMA_CHARS = 10000
if preferred_provider == "ollama" and len(context_str) > MAX_OLLAMA_CHARS:
logger.warning(f"⚠️ [{query_id}] Context zu groß für Ollama ({len(context_str)} chars). Kürze auf {MAX_OLLAMA_CHARS}.") settings = get_settings() # Falls noch nicht im Scope vorhanden
context_str = context_str[:MAX_OLLAMA_CHARS] + "\n[...gekürzt zur Stabilität...]" max_chars = getattr(settings, "MAX_OLLAMA_CHARS", 10000)
if preferred_provider == "ollama" and len(context_str) > max_chars:
logger.warning(f"⚠️ [{query_id}] Context zu groß für Ollama ({len(context_str)} chars). Kürze auf {max_chars}.")
context_str = context_str[:max_chars] + "\n[...gekürzt zur Stabilität...]"
template = llm.get_prompt(prompt_key) or "{context_str}\n\n{query}" template = llm.get_prompt(prompt_key) or "{context_str}\n\n{query}"