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.

Sign up for GitHub

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

Jump to bottom

[Inductor][Flex-attention] Support different sequence lengths for Query and Key/Value #126639

Closed
wants to merge 2 commits into from
Closed

[Inductor][Flex-attention] Support different sequence lengths for Query and Key/Value #126639

wants to merge 2 commits into from

Conversation

yanboliang
Copy link
Contributor

@yanboliang yanboliang commented May 19, 2024 •
edited by pytorch-bot bot

@pytorch-bot PyTorch Bot
Copy link

pytorch-bot bot commented May 19, 2024 •
edited

🔗 Helpful Links

🧪 See artifacts and rendered test results at hud.pytorch.org/pr/126639

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

❌ 1 New Failure, 6 Unrelated Failures

As of commit 3678fe5 with merge base 41fb4bc (image):

NEW FAILURE - The following job has failed:

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

UNSTABLE - The following jobs failed but were likely due to flakiness present on trunk and has been marked as unstable:

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

@yanboliang yanboliang marked this pull request as ready for review May 19, 2024 04:56
Chillee
Chillee reviewed May 19, 2024
View reviewed changes
test/inductor/test_flex_attention.py Outdated
self._check_equal(golden_out, ref_out, compiled_out, fudge_factor, "Out")

# Check gradients
q_fudge_factor = 2.5 * fudge_factor
Copy link
Contributor

Choose a reason for hiding this comment

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

Can we refactor this a bit lol. I think we have this code copied 3 different times in this file.

torch/_inductor/kernel/flex_attention.py Outdated
@@ -188,34 +188,36 @@ def build_subgraph_buffer(

Z = {{size("Q", 0)}}
H = {{size("Q", 1)}}
N_CTX = {{size("Q", 2)}}
M = {{size("Q", 2)}}
Copy link
Contributor

Choose a reason for hiding this comment

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

Let's call this Q_LEN and KV_LEN?

@albanD albanD removed their request for review May 20, 2024 14:31
@yanboliang
Copy link
Contributor Author

closing this in favor of #127678

@yanboliang yanboliang closed this Jun 1, 2024
@yanboliang yanboliang deleted the flex-diff-q branch June 1, 2024 05:02
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Reviewers

@Chillee Chillee Chillee left review comments

@jbschlosser jbschlosser Awaiting requested review from jbschlosser jbschlosser is a code owner

@mikaylagawarecki mikaylagawarecki Awaiting requested review from mikaylagawarecki mikaylagawarecki is a code owner

@drisspg drisspg Awaiting requested review from drisspg

Assignees
No one assigned
Labels
ciflow/inductor module: inductor
Projects
None yet
Milestone
No milestone
Development

Successfully merging this pull request may close these issues.

None yet
2 participants
@yanboliang @Chillee