[schmeichel]

I'd recommended using llama.cpp and The Bloke's GGUF version of this model!

https://github.com/ggerganov/llama.cpp/ https://huggingface.co/TheBloke/MonadGPT-GGUF