config: model profiles inherit the standard port/container

Replace the stale example profiles (qwen35->8002, deepseek->8003, pointing at
non-existent paths) with profiles for the actually installed models. They now
override ONLY model_path and inherit host_port=8001, container_name and alias
from [default] -> one model at a time on the standard port; --start/--change
swaps it. Distinct port/container is only needed for concurrent operation.

Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>

llama.cpp.config

@@ -57,22 +57,20 @@ max_tokens = 32768
 # chat_temperature leer lassen -> Server-Temp (temp oben) gilt.
 chat_temperature =
 
-# Example second model profile, pinned to the second GPU (e.g. RTX 3090 #2)
+# Modell-Profile: erben ALLES aus [default] (Port 8001, Container
-# with its own port and container name so it can run alongside [default].
+# llama_cpp_server, Alias default_llm, gpu_device 1, ctx_size 262144) und
-[model.qwen35]
+# überschreiben NUR das Modell. Es läuft also immer ein Modell auf dem
-model_path = qwen3/Qwen3-35B-A3B-Q4_K_M.gguf
+# Standard-Port; --start/--change tauscht es aus. Wer Modelle GLEICHZEITIG
-model_alias = qwen35
+# betreiben will, gibt dem Profil zusätzlich einen eigenen container_name +
-container_name = llama_cpp_qwen35
+# host_port (Muster siehe llama.cpp.config.example).
-host_port = 8002
+[model.carnice]
-gpu_device = 1
+model_path = models/qwen3/Carnice-Qwen3.6-MoE-35B-A3B-Q4_K_M.gguf
 
-[model.deepseek]
+[model.qwen27]
-model_path = deepseek/deepseek-r1-q4.gguf
+model_path = models/qwen3/Qwen3.6-27B-Uncensored-HauhauCS-Aggressive-IQ4_XS.gguf
-model_alias = deepseek
+
-container_name = llama_cpp_deepseek
+[model.qwopus]
-host_port = 8003
+model_path = models/qwen3/Qwopus3.6-35B-A3B-v1-Q4_K_M.gguf
-gpu_device = 1
-ctx_size = 131072
 
 [prompt.concise]
 system_prompt = Du antwortest kurz, präzise und technisch.