cturan/llama.cpp @ e81b8e4b7f5ab870836fad26d154a7507b341b36

의 미러 https://github.com/cturan/llama.cpp

Johannes Gäßler e81b8e4b7f llama: use FA + max. GPU layers by default (#15434)		4 달 전
..
llama-cpp.h	afa8a9ec9b llama : add `llama_vocab`, functions -> methods, naming (#11110)	1 년 전
llama.h	e81b8e4b7f llama: use FA + max. GPU layers by default (#15434)	4 달 전