[torch.compile] Adding torch compile annotations to some models

[CRZbulabula]

This PR is to improve #9589, adding torch compile annotations to some models and testing them.

[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.

🚀

[youkaichao]

this will actually cause on error, but let me merge it first, so that pytorch team can easily reproduce it.

[youkaichao]

the error is:

RuntimeError: aten/src/ATen/RegisterCUDA.cpp:7275: SymIntArrayRef expected to contain only concrete integers

[youkaichao]

thanks for the great work!