PSA: Phi-4 GGUFs (from official release) don't work in Ollama (yet)

I assumed no changes were needed (since it's the same arch) but there were changes made here in llama.cpp that haven't been picked up by them yet (they are aware): https://github.com/ggerganov/llama.cpp/pull/10817/files In the meantime LM Studio is confirmed working, and Koboldcpp seems to have picked up the changes as well I can't tell if text-gen-webui would work, it's using llama-cpp-python 0.3.5 which is 4 weeks old where the llama.cpp changes were made 3 weeks ago, so I assume that it wouldn't work (unless it has its own workaround)