Hacker News new | ask | show | jobs
by zozbot234 80 days ago
KV quantization has long been available in llama.cpp
1 comments
chr15m 79 days ago
Yes but the optimisation described has not right?
link