How to Use Llama.cpp to Execute Prompts from Text Files

[smy20011]

Hey, I'm trying to run a local LLM pipeline:

1. I have a list of instructions in ".md" files.
2. I want to run Llama.cpp in chat completion mode, which reads these .md files and performs a single chat completion for each.
3. The chat completion output should be written to a file.

It seems to work well if I paste the content of an .md file directly into the chat mode, but I'm having trouble using the -f flag to read the prompt from a file. The -f command replaces the entire prompt and doesn't stop properly.

Any help is appreciated!