llama.cpp

9ca2e677 - server : add speculative decoding support (#10455)

server : add speculative decoding support (#10455) * server : add speculative decoding support ggml-ci * server : add helper function slot.can_speculate() ggml-ci