noerms on Nostr: I successfully installed llama.cpp and Qwen3 8B Q4_K_M. Even docker and Open WebUI ...

I successfully installed llama.cpp and Qwen3 8B Q4_K_M. Even docker and Open WebUI were installed without any problems. Conbected open webui to qwen3. But now every time i start the mist simple promt like „just say hello“ it starts reasoning like forever. /no_think or custom parameters wont help. Do you maybe have any idea how i can stop the thinking ?