[AMD][ROCm] Enable DeepSeek model on ROCm

[hongxiayang]

Thanks for great work from vLLM community to enable DeepSeek model.
This PR is to use the Triton MLA backend on ROCm to enable DeepSeek model on AMD GPUs.

Tests:

• Tested on DeepSeek V2 (DeepSeek-Coder-V2-Lite-Instruct) model on MI300x and MI210.
• Tested on DeepSeek V3 dummy weights on MI300x.
• Added two unit tests.

Note: to use DeepSeek model, we should ensure to have the "trust_remote_code" flag on.

[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.

🚀

[mgoin]

LGTM! Would you able to add DeepSeek-Coder-V2-Lite-Instruct as a model test for the AMD runner with a check that TritonMLA is chosen for the attention module?

[mgoin]

Thank you

[hongxiayang]

LGTM! Would you able to add DeepSeek-Coder-V2-Lite-Instruct as a model test for the AMD runner with a check that TritonMLA is chosen for the attention module?

@mgoin

Thank you

@mgoin Thank your for your review and approval.