Skip to content

Set attn backend to eager for pooling models#1028

Merged
maxdebayser merged 2 commits into
torch-spyre:mainfrom
maxdebayser:fix_encoder_eager_attn
Jul 10, 2026
Merged

Set attn backend to eager for pooling models#1028
maxdebayser merged 2 commits into
torch-spyre:mainfrom
maxdebayser:fix_encoder_eager_attn

Conversation

@maxdebayser

Copy link
Copy Markdown
Collaborator

With transformers 5 the default became "sdpa" and that is causing very long compilation times as well as
compiler crashes

With transformers 5 the default became "sdpa" and that
is causing very long compilation times as well as
compiler crashes

Signed-off-by: Max de Bayser <mbayser@br.ibm.com>
@github-actions

Copy link
Copy Markdown

👋 Hi! Thank you for contributing.
Just a reminder: Make sure that your code passes all the linting checks, otherwise your PR won't be able to be merged. To do so, run ./format.sh.
Now you are good to go 🚀.

We also recommend installing prek and configuring it to check your code before every local commit.

@gkumbhat gkumbhat left a comment

Copy link
Copy Markdown
Collaborator

Choose a reason for hiding this comment

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

Looks good to me!

Signed-off-by: Max de Bayser <mbayser@br.ibm.com>
@maxdebayser maxdebayser enabled auto-merge (squash) July 10, 2026 20:11
@github-actions github-actions Bot added the ready Runs the full CI test suite. Only add to PRs once ready to merge to limit public GHA usage label Jul 10, 2026
@maxdebayser maxdebayser merged commit 8baaae8 into torch-spyre:main Jul 10, 2026
18 of 24 checks passed
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

ready Runs the full CI test suite. Only add to PRs once ready to merge to limit public GHA usage

Projects

None yet

Development

Successfully merging this pull request may close these issues.

2 participants