Please add multi token prediction https://github.com/ggml-org/llama.cpp/pull/24282
Please add multi token prediction
ggml-org/llama.cpp#24282