What are your run params? By the way, posts such as these are worth their weight in gold :D Thanks!

I just tried that exact prompt with:

  "MiniMax-M27-IQ4":
...
      docker run --rm --init --label llama-swap.managed=true
      --name llama-cpp-minimax-m27-iq4
      --gpus all
      --ipc=host
      -p ${PORT}:8080
      -v /home/user/prj/llama-swap/models/.cache/llama.cpp:/root/.cache/llama.cpp
      -v /home/user/prj/llama-swap/models:/models
      llama-swap/llama-cpp:b8775
      --model /root/.cache/llama.cpp/MiniMax-M2.7-UD-IQ4_XS/UD-IQ4_XS/MiniMax-M2.7-UD-IQ4_XS-00001-of-00004.gguf
      --host 0.0.0.0
      --port 8080
      --ctx-size 132000
      --parallel 1
      --jinja
      --temp 1.0
      --top-p 0.95
      --top-k 40
      --repeat-penalty 1.05

And it was worse. I had to add repeat penalty, because it looped like crazy otherwise