llama-bench

参考

• GitHub: ggml-org/llama.cpp/examples/llama-bench/README.md

• AMD GPU support for llama.cpp via Vulkan on Raspberry Pi 5

• How to benchmark 'llama.cpp' builds for specific hardware?

• Performance of llama.cpp on Apple Silicon M-series #4167 提供了如何对比测试的案例