Skip to content

Conversation

@Alcpz
Copy link
Contributor

@Alcpz Alcpz commented Nov 18, 2024


Reverts the changes introduced in #10257 that introduced a performance regression as it disabled more MUL_MAT operations than intended.

Be wary that these changes will make the SYCL test-backend-ops to fail.

@github-actions github-actions bot added the SYCL https://en.wikipedia.org/wiki/SYCL - GPU programming language label Nov 18, 2024
@Alcpz
Copy link
Contributor Author

Alcpz commented Nov 18, 2024

@NeoZhangJianyu reverting here. There's still an issue with quantizations that use get_rows, but that is better addressed in a different PR I think, as it was introduced in a different place.

@NeoZhangJianyu NeoZhangJianyu merged commit 557924f into ggml-org:master Nov 19, 2024
54 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

SYCL https://en.wikipedia.org/wiki/SYCL - GPU programming language

Projects

None yet

Development

Successfully merging this pull request may close these issues.

2 participants