Replies: 1 comment
-
@dubefab I think you could launch a test for different building settings. For instance, I have tried to build with CoreML and CoreML + Embed the Metal Library. I use While it is not a serious or big benchmark test, based on the result, I found that
Just for reference. |
Beta Was this translation helpful? Give feedback.
0 replies
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
-
Hey! I'm pretty satisfied with whisper.cpp leveraging Metal on my m1 pro.
I am curious: if I build and use the CoreML model: will it speed it up even further?
Or is one or the other (pick between Metal or CoreML)? Not together?
Beta Was this translation helpful? Give feedback.
All reactions