Skip to content
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

[TIR][Schedule] Enhance compute-inline for fusion #15142

Merged

Conversation

junrushao
Copy link
Member

This PR enhances compute-inline and reverse-compute-inline to handle more complicated fusion patterns where the RHS of the equation could be generic expressions rather than strict buffer indexing. This could be used in multiple cases in LLM inference.

@tvm-bot
Copy link
Collaborator

tvm-bot commented Jun 22, 2023

Thanks for contributing to TVM! Please refer to the contributing guidelines https://tvm.apache.org/docs/contribute/ for useful information and tips. Please request code reviews from Reviewers by @-ing them in a comment.

Generated by tvm-bot

@junrushao junrushao marked this pull request as ready for review June 22, 2023 05:10
@junrushao junrushao force-pushed the feature/2023-06-21/upstream-compute-inline branch 5 times, most recently from 0bc3eed to ae42a89 Compare June 24, 2023 05:24
@junrushao
Copy link
Member Author

I just realized that tests/python/unittest/test_tir_schedule_tensorize_ldmatrix_mma.py::test_i8i8i32_m16n16k32 fails on my machine when running the tests. It's likely not tested on CI environment

@junrushao
Copy link
Member Author

@tqchen @Hzfengsy This PR should be good to go

Copy link
Member

@Hzfengsy Hzfengsy left a comment

Choose a reason for hiding this comment

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

Please address the sm version comment, and we can get it in

This PR enhances `compute-inline` and `reverse-compute-inline` to handle
more complicated fusion patterns where the RHS of the equation could be
generic expressions rather than strict buffer indexing. This could be
used in multiple cases in LLM inference.
@junrushao junrushao force-pushed the feature/2023-06-21/upstream-compute-inline branch from ae42a89 to dd07026 Compare June 24, 2023 18:40
@tqchen tqchen merged commit 5a3523d into apache:main Jun 24, 2023
5 checks passed
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.

None yet

4 participants