Skip to content

Conversation

@vinx13
Copy link
Member

@vinx13 vinx13 commented Apr 13, 2024

This makes weight loading from global to local vectorized. Usually for packed weight in uint32 format there is not much difference since uint32 already implies vectorization. This PR enhances vectorization when the weight is not packed.

cc @spectrometerHBH @tqchen

@tqchen tqchen merged commit 5c80691 into apache:main Apr 13, 2024
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

2 participants