You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
I wonder whether GGUF ALWAYS stores the weights transposed.
To make it clear, the physical storage layout might be different to its logical shape.
Why? Does this layout designed to optimized for a specific architecture like GPU?
I find that it hurts the cache performance on my target architecture.
Is there a tool to convert the storage layout back to non-transposed?
reacted with thumbs up emoji reacted with thumbs down emoji reacted with laugh emoji reacted with hooray emoji reacted with confused emoji reacted with heart emoji reacted with rocket emoji reacted with eyes emoji
-
I wonder whether GGUF ALWAYS stores the weights transposed.
To make it clear, the physical storage layout might be different to its logical shape.
Why? Does this layout designed to optimized for a specific architecture like GPU?
I find that it hurts the cache performance on my target architecture.
Is there a tool to convert the storage layout back to non-transposed?
Beta Was this translation helpful? Give feedback.
All reactions