llama.cpp icon indicating copy to clipboard operation
llama.cpp copied to clipboard

Too slow on m2 MBA 16gb SSD 512GB

Open effortprogrammer opened this issue 1 year ago • 3 comments

Hi,

First of all, thanks for the tremendous work!

I just wanted to ask that compared to your demo, when I run the same input sentence, the speed difference is tremendously different. Is this because of the chipset difference between m1 pro and m2 or, you already knew this issue and trying to fix this?

effortprogrammer avatar Mar 11 '23 22:03 effortprogrammer

How much slower? Post the stats from the end of the run with the model used

Lupul avatar Mar 12 '23 02:03 Lupul

Also post the command-line you are using

prusnak avatar Mar 12 '23 10:03 prusnak

try ./main -h find -t args means number of threads to use, which would make it really fast, like 0.2s per token.

shm007g avatar Mar 21 '23 09:03 shm007g

No reaction from the reporter. Please reopen if the issue still persists.

prusnak avatar Apr 16 '23 09:04 prusnak