[tools]
"llama.cpp" = { version = "latest", prerelease = "true", version_prefix = "b" }

[tasks.serve]
run = "llama serve --model ./llm/model2.gguf --port 8080 -ngl 99 --ctx-size 16384"