ggml_backend_metal_set_n_cb: n_cb = 24, using n_cb > 2 is not recommended

[CantHackTheHack]

Hello. I have been spending a lot of time trying to tune a moe model to get the max response time out of it. This line keeps showing up in the logs once a prompt is initiated. I see where the log is set in source, tried to for to 2 without much success so far. Any thoughts with this error and how to handle it? Thank you.