[Bugfix][Misc]: fix graph capture for decoder

[yudian0504]

This PR fixes a bug introduced by #8645.

related issues: #9448 & #9417 & #9409

[github-actions]

👋 Hi! Thank you for contributing to the vLLM project.
Just a reminder: PRs would not trigger full CI run by default. Instead, it would only run fastcheck CI which starts running only a small and essential subset of CI tests to quickly catch errors. You can run other CI tests on top of those by going to your fastcheck build on Buildkite UI (linked in the PR checks section) and unblock them. If you do not have permission to unblock, ping simon-mo or khluu to add you in our Buildkite org.

Once the PR is approved and ready to go, your PR reviewer(s) can run CI to test the changes comprehensively before merging.

To run CI, PR reviewers can do one of these:

• Add ready label to the PR
• Enable auto-merge.

🚀

[comaniac]

Approved the unblock. Meanwhile, I'm not sure how encoder-decoder works in this case. @varun-sundar-rabindranath PTAL. Also cc @sroy745

[varun-sundar-rabindranath]

My mistake !! Thanks for the fix @yudian0504 🙌

[jeejeelee]

Good catch, this solved a bug I encountered.

[xfalcox]

This bug fix is so important, we need a release for the new docker image cc @DarkLight1337

[DarkLight1337]

This bug fix is so important, we need a release for the new docker image cc @DarkLight1337

For now, you can use the Docker image here with the latest commit.