dlxj
todo: 基于 CUDA 13.0 编译
2517be1
|
Raw
History Blame Contribute Delete
260 Bytes

llama.cpp/example/simple-chat

The purpose of this example is to demonstrate a minimal usage of llama.cpp to create a simple chat program using the chat template from the GGUF file.

./llama-simple-chat -m Meta-Llama-3.1-8B-Instruct.gguf -c 2048
...