mirror of
https://github.com/ikawrakow/ik_llama.cpp.git
synced 2026-02-25 23:54:10 +00:00
PP is not too bad - just 10% slower than q4_0. But TG is 30% slower, i.e., predictably bad.
PP is not too bad - just 10% slower than q4_0. But TG is 30% slower, i.e., predictably bad.