llama.cpp

c5fef0fc - server: update readme to mention n_past_max metric (#16436)

server: update readme to mention n_past_max metric (#16436) https://github.com/ggml-org/llama.cpp/pull/15361 added new metric exported, but I've missed this doc.