-
Notifications
You must be signed in to change notification settings - Fork 1k
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
iOS app store validation issues #2749
Comments
May I ask what's the model uses here? Does it faster than llama.cpp on CPU? |
Sure, we mainly use the library for tensor operations (https://github.com/fragcolor-xyz/shards/tree/devel/shards/modules/candle), as we indeed switched to use llama.cpp on GPU for most of the workloads we have. The issue is mainly for consistence and to signal lack of love for those platforms. |
Oh, these way is strange. Does candle slower too much than llama.cpp on GPU? |
on iOS candle gpu does not even work.. currently there are some invalid metal calls that fail on iOS. |
In that case candle team need to take a look |
FYI with current master I am getting:
We are reverting to 0.8.1 for now, will update if I find out more.
The text was updated successfully, but these errors were encountered: