Bug: template parsing issues with Apertus

### Contact Details _No response_ ### What happened? If I run llamafile as `~/workspace/llamafile/o/llamafile/llamafile -m Apertus-8B-Instruct-2509-Q6_K.gguf --server` I get the following: ``` Hello, who are you? Hello! My name is Apertus, and I am an AI assistant developed by the SwissAI initiative. I am here to help you with any questions or information you might need within my knowledge cutoff date of 2024. How can I assist you today? ``` If, instead, I run llamafile from the TUI, as `~/workspace/llamafile/o/llamafile/llamafile -m Apertus-8B-Instruct-2509-Q6_K.gguf`, I get: ``` >>> Hello, who are you? Вопрос: который вы задали, не совсем понятен. Однако, я могу предположить, что вы имеете в виду, как можно использовать искусственный интеллект ``` 😬 It looks like Apertus has its own [template](https://github.com/ggml-org/llama.cpp/issues/15748), my hypothesis is that we are just not loading it properly ### Version v0.10.0 alpha ### What operating system are you seeing the problem on? Mac ### Relevant log output ```shell ```