llama.cpp
llama.cpp copied to clipboard
Published
20 hours ago •
ggml-org
ggml-org
Using OpenBlas to accelerate has no effect？
When compiling, openblas was enabled, but it seems that there is no acceleration effect during inference. Compared to not enabling openblas, it only increases the memory usage. What is the reason for this? How much faster can OpenBlas accelerate under normal circumstances?
