PR #3115 llama : make quantize example up to 2.7x faster

llama : refactor k-quant mixture logic into a function

cebtenzzre committed 2 years ago

llama : optimize vector use in quantize -> 179% faster

cebtenzzre committed 2 years ago

llama : don't zero-init vectors in quantize -> 5.1% faster

cebtenzzre committed 2 years ago

llama.cpp llama : make quantize example up to 2.7x faster #3115 Merged