Skip to content

[ROCm] Fix Sliding Window Attention in AOTriton integration code #159773

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

Draft
wants to merge 2 commits into
base: main
Choose a base branch
from

Conversation

xinyazhang
Copy link
Collaborator

@xinyazhang xinyazhang commented Aug 4, 2025

AOTriton implements Sliding Window Attention (SWA) as a more generalized version of causal masks and also needs an atomic counter for dynamic workload allocation.

Fixes #158308

cc @jbschlosser @jeffdaily @sunway513 @jithunnair-amd @pruthvistony @ROCmSupport @dllehr-amd @jataylo @hongxiayang @naromero77amd

Copy link

pytorch-bot bot commented Aug 4, 2025

🔗 Helpful Links

🧪 See artifacts and rendered test results at hud.pytorch.org/pr/159773

Note: Links to docs will display an error until the docs builds have been completed.

❌ 4 Cancelled Jobs, 3 Unrelated Failures

As of commit 759a982 with merge base bb67660 (image):

CANCELLED JOBS - The following jobs were cancelled. Please retry:

FLAKY - The following jobs failed but were likely due to flakiness present on trunk:

UNSTABLE - The following job is marked as unstable, possibly due to flakiness on trunk:

This comment was automatically generated by Dr. CI and updates every 15 minutes.

@xinyazhang
Copy link
Collaborator Author

@pytorchbot label "topic: not user facing"

@pytorch-bot pytorch-bot bot added the topic: not user facing topic category label Aug 4, 2025
@xinyazhang xinyazhang added the module: cpp Related to C++ API label Aug 5, 2025
@xinyazhang
Copy link
Collaborator Author

@jithunnair-amd we need to add label to describe C++ API changes for PyTorch 2.9 and remove the "topic: not user facing" label

@pruthvistony pruthvistony added ciflow/rocm Trigger "default" config CI on ROCm ciflow/inductor-rocm Trigger "inductor" config CI on ROCm ciflow/rocm-mi300 Trigger "default" config CI on ROCm MI300 ciflow/periodic-rocm-mi300 Trigger "distributed" config CI on ROCm MI300 labels Aug 8, 2025
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
ciflow/inductor-rocm Trigger "inductor" config CI on ROCm ciflow/periodic-rocm-mi300 Trigger "distributed" config CI on ROCm MI300 ciflow/rocm Trigger "default" config CI on ROCm ciflow/rocm-mi300 Trigger "default" config CI on ROCm MI300 module: cpp Related to C++ API module: rocm AMD GPU support for Pytorch open source topic: not user facing topic category
Projects
None yet
Development

Successfully merging this pull request may close these issues.

Memory Error for _flash_attention_forward on AMD Radeon RX 7900 XTX
3 participants