Skip to content

Conversation

jeffbolznv
Copy link
Collaborator

Integer div/mod is relatively expensive and not needed for the most common case (batch_idx 0). Perf gain is very small (in the noise) and I don't want to misrepresent it by posting numbers, but it helps all formats by some epsilon amount and it's an easy improvement.

@jeffbolznv jeffbolznv requested a review from 0cc4m November 26, 2024 00:43
Copy link
Collaborator

@0cc4m 0cc4m left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Looks good, thank you.

@0cc4m 0cc4m merged commit 71a6498 into ggml-org:master Nov 27, 2024
7 checks passed
arthw pushed a commit to arthw/llama.cpp that referenced this pull request Dec 20, 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