"I'd like to see other models."

[BradHutchings]

A note that might be helpful to those doing local installations.

You need to find models that have been converted to GGUF format. That's the newish format supported by llama.cpp. It replaces GGML, which is no longer supported by llama.cpp.

I wanted to try out Apple's OpenELM. It turns out, they don't support GGUF out of the box. There is a lonely github developer working on it.

ggerganov/llama.cpp#6868

All that aside, I have worked out the settings-local.yaml file. a small change to scripts/setup, and the order of operations to use a different GGUF model. I'll post again in the thread when I find something suitable.