/workspace/llama-server -m /workspace/model/Qwen3.6-27B-Omnimerge-v4-Q4_K_M.gguf --host 0.0.0.0 --port 5000 -ngl 99 -t 8 --parallel 2 -c 262144 --mlock  --cont-batching --reasoning off -ctk q8_0  -ctv q8_0 --no-context-shift