[ROCm] [CI] Add More Tests

[tjtanaa]

PLEASE FILL IN THE PR DESCRIPTION HERE ENSURING ALL CHECKLIST ITEMS (AT THE BOTTOM) HAVE BEEN CONSIDERED.

Purpose

Adding four more tests. Reaching test parity as on CUDA.

Tests added:

• Diffusion Parallelism Test
• Diffusion GPU Worker Test
• Omni Model Test Qwen3-Omni (supports CUDA Graph)
• Diffusion Image Edit Test

It also addressed the ECR repository rate limit.

Test Plan

Pass the CI

Test Result

CI results https://buildkite.com/vllm/vllm-omni-amd-ci/builds/301/steps/canvas

---

Essential Elements of an Effective PR Description Checklist

• The purpose of the PR, such as "Fix some issue (link existing issues this PR will resolve)".
• The test plan, such as providing test command.
• The test results, such as pasting the results comparison before and after, or e2e results
• (Optional) The necessary documentation update, such as updating supported_models.md and examples for a new model.
• (Optional) Release notes update. If your change is user facing, please update the release notes draft.

BEFORE SUBMITTING, PLEASE READ https://github.com/vllm-project/vllm-omni/blob/main/CONTRIBUTING.md (anything written below this line will be removed by GitHub Actions)

[tjtanaa]

@hsliuustc0106 it is ready for review. I have also synced with main. The Qwen3 unit tests runs with Cudagraph now.

[tjtanaa]

@hsliuustc0106 @ywang96 Ready for review.

[hsliuustc0106]

shall we wait for #577

[tjtanaa]

@hsliuustc0106 can we merge this first so that we test more cases because that PR is still work-in-progress?

Or is there a timeline when will that PR be ready?

[hsliuustc0106]

@hsliuustc0106 can we merge this first so that we test more cases because that PR is still work-in-progress?

Or is there a timeline when will that PR be ready?

OK, let's try to merge it first

[tjtanaa]

@hsliuustc0106 all check passed.

[hsliuustc0106]

got you

[hsliuustc0106]

lgtm