Model updates ask before downloading, E2B and E4B default to Q6_K
Performance: LLAMACPP_PARALLEL
llama.cpp is the default provider: install steps, logs, model swap, where she landed, AMD and uninstall say so, Ollama stays as the Custom pick
JunOS wiki Co-Authored-By: MkJez-R <309192026+MkJez-R@users.noreply.github.com>