How is the GGUF file used by the llama-server to produce the ouput?

[yogheswaran-A]

Hi,
I am trying to learn how the llama.cpp repo works. I have quantized llama 3.1 8B and stored the data in GGUF format as described in the llama.cpp repo. I want to know which script(s) is(are) responsible for using the GGUF file to produce the outputs for the input tokens when llama-server is called.

[ag2s20150909]

You can refer to the following

https://github.com/ggml-org/llama.cpp/tree/master/examples/simple

https://github.com/ggml-org/llama.cpp/tree/master/examples/simple-chat

[yogheswaran-A]

@ag2s20150909 Thanks a ton!! It seems like this what I was looking for.